SO2) National Ambient Air Quality Standards (NAAQS

Attachment 1

Indiana’s Preliminary Recommendations Concerning Round 3 Designations for the 2010 Primary 1-Hour Sulfur Dioxide (SO 2) National Ambient Air Quality Standards (NAAQS) This page left intentionally blank. Source County Boundary/Area Recommendation

Duke Energy Gallagher Floyd County Attainment U.S. Mineral Products (Isolatek) Huntington See Attachment 3 See Attachment 3 NIPSCO R.M. Schahfer Jasper (P) Kankakee Township Attainment ArcelorMittal USA, Coke Energy, Calumet Township and Lake (P) Attainment U.S. Steel Gary Works North Township SABIC Innovative Plastics Posey (P) Black Township Attainment Hoosier Energy Merom Sullivan (P) Gill Township Attainment Eugene Township Duke Energy Cayuga Vermillion (P) Attainment Vermillion Township Alcoa Warrick Power Plant, Attainment Warrick (P) Anderson Township Alcoa Warrick Operations Plant See Attachment 4 (P) denotes partial county recommendation

The following sources, though identified by U.S. EPA as being subject to the Data Requirements Rule, were addressed under Round 2 designations and designated by U.S. EPA on June 30, 2016 (81 FR 45039).

Area Name Source Area County Name Designation

Gibson County, IN Duke Gibson Gibson Unclassifiable/Attainment Jefferson County, IN IKEC Clifty Creek Jefferson (P) Unclassifiable/Attainment LaPorte County, IN NIPSCO Michigan City LaPorte Unclassifiable/Attainment Posey County, IN Vectren A.B. Brown Posey (P) Unclassifiable/Attainment Spencer County, IN AEP Rockport Spencer (P) Unclassifiable/Attainment

Page 1 of 1

This page left intentionally blank. Attachment 2

Indiana’s Air Quality Modeling Technical Support Document

Preliminary Designation Recommendations

Data Requirements Rule (Round 3)

2010 Primary 1-Hour Sulfur Dioxide

(SO 2) National Ambient Air Quality Standard (NAAQS)

January 2017

Table of Contents

1.0 1-Hour SO 2 NAAQS and Designation Process ...... 1 2.0 Data Requirements Rule ...... 2 3.0 Methodology for DRR Air Quality Modeling ...... 4 4.0 Model Selection for DRR Modeling ...... 5 4.1 AERMOD Dispersion Model ...... 5 4.2 AERMAP ...... 5 4.3 Land Use Determination ...... 5 5.0 Receptor Grid and Modeling Domain ...... 5 6.0 Meteorology ...... 6 6.1 AERMET ...... 6 6.2 AERMINUTE/AERSURFACE ...... 7

7.0 SO 2 Background Concentrations ...... 7

8.0 SO 2 Emissions Sources to be Modeled ...... 8 8.1 DRR Sources ...... 8 8.2 Inventory Sources ...... 9 8.3 Intermittent Sources ...... 9 9.0 Analysis of Modeling Results ...... 10 10.0 ArcelorMittal – Burns Harbor (Source ID 18-127-00001) ...... 10 10.1 Source Description ...... 10 10.2 Characterization of Modeled Area ...... 10 10.3 Summary of DRR Monitoring Approach ...... 11 11.0 SABIC Innovative Plastics (Source ID 18-129-00002) ...... 11 11.1 Source Description ...... 11 11.2 Characterization of Modeled Area ...... 12 11.3 Background Concentrations ...... 12 11.4 Modeling Methodology ...... 13 11.4.1 Model Selection ...... 13 11.4.2 Model Options ...... 13

11.4.3 AERMAP ...... 14 11.5 Meteorological Data ...... 14 11.5.1 AERMET ...... 14 11.5.2 Wind Rose ...... 15 11.5.3 AERMINUTE/AERSURFACE ...... 15 11.6 Receptor Grid and Modeling Domain ...... 16 11.7 Stack Heights ...... 17

11.8 Temporally Varying Seasonal 1-Hour SO 2 Background ...... 17

11.9 SO 2 Emissions Included in the Modeling Analysis ...... 18 11.9.1 DRR Source: SABIC Emissions ...... 18

11.9.2 Inventoried SO 2 Sources Included in the Modeling ...... 19 11.10 Modeling Results ...... 20 12.0 Lake County: Source IDs ArcelorMittal – USA (18-089-00316)/Cokenergy (18-089-00383)/U.S. Steel (18-089-00121) ...... 22 12.1 Source Description ...... 22 12.2 Characterization of Modeled Area ...... 22 12.3 Background Concentrations ...... 23 12.4 Modeling Methodology ...... 24 12.4.1 Model Selection ...... 24 12.4.2 Model Options ...... 24 12.4.3 AERMAP ...... 25 12.5 Meteorological Data ...... 25 12.5.1 AERMET ...... 25 12.5.2 Wind Rose ...... 26 12.5.3 AERMINUTE/AERSURFACE ...... 26 12.6 Receptor Grid and Modeling Domain ...... 27 12.7 Stack Heights ...... 28

12.8 Temporally Varying Seasonal 1-Hour SO 2 Background ...... 28

12.9 SO 2 Emissions Included in the Modeling Analysis ...... 30 12.9.1 DRR Source Emissions ...... 30

12.9.2 Carmeuse Lime’s Commissioner’s Order – SO 2 Emission Limits ...... 30

12.9.3 Inventoried SO 2 Sources Included in the Modeling ...... 31 12.10 Modeling Results ...... 32 13.0 Duke-Gallagher (Source ID 153-00005) ...... 34 13.1 Source Description ...... 34 13.2 Characterization of Modeled Area ...... 34 13.3 Background Concentrations ...... 35 13.4 Modeling Methodology ...... 35 13.4.1 Model Selection ...... 35 13.4.2 Model Options ...... 35 13.4.3 AERMAP ...... 36 13.5 Meteorological Data ...... 36 13.5.1 AERMET ...... 36 13.5.2 Wind Rose ...... 37 13.5.3 AERMINUTE/AERSURFACE ...... 37 13.6 Receptor Grid and Modeling Domain ...... 38 13.7 Stack Heights ...... 39

13.8 Temporally Varying Seasonal 1-Hour SO 2 Background ...... 39

13.9 SO 2 Emissions Included in the Modeling Analysis ...... 40 13.9.1 DRR Source: Duke - Gallagher Emissions ...... 40

13.9.2 Inventoried SO 2 Sources Included in the Modeling ...... 41 13.10 Modeling Results ...... 41 14.0 NIPSCO – R.M. Schahfer Generating Station (Source ID 18-073-00008) ...... 43 14.1 Source Description ...... 43 14.2 Characterization of Modeled Area ...... 43 14.3 Background Concentrations ...... 44 14.4 Modeling Methodology ...... 44 14.4.1 Model Selection ...... 44 14.4.2 Model Options ...... 44 14.4.3 AERMAP ...... 45 14.5 Meteorological Data ...... 45 14.5.1 AERMET ...... 45 iii

14.5.2 Wind Rose ...... 46 14.5.3 AERMINUTE/AERSURFACE ...... 46 14.6 Receptor Grid and Modeling Domain ...... 47 14.7 Stack Heights ...... 48

14.8 Temporally Varying Seasonal 1-Hour SO 2 Background ...... 48

14.9 SO 2 Emissions Included in the Modeling Analysis ...... 49 14.9.1 DRR Source: NIPSCO - Schahfer Generating Station Emissions ...... 49

14.9.2 Inventoried SO 2 Sources Included in the Modeling ...... 50 14.10 Modeling Results ...... 50 15.0 Hoosier Energy - Merom (Source ID 153-00005) ...... 51 15.1 Source Description ...... 51 15.2 Characterization of Modeled Area ...... 52 15.3 Background Concentrations ...... 52 15.4 Modeling Methodology ...... 53 15.4.1 Model Selection ...... 53 15.4.2 Model Options ...... 53 15.4.3 AERMAP ...... 54 15.5 Meteorological Data ...... 54 15.5.1 AERMET ...... 54 15.5.2 Wind Rose ...... 55 15.5.3 AERMINUTE/AERSURFACE ...... 55 15.6 Receptor Grid and Modeling Domain ...... 56 15.7 Stack Heights ...... 57

15.8 Temporally Varying Seasonal 1-Hour SO 2 Background ...... 57

15.9 SO 2 Emissions Included in the Modeling Analysis ...... 58 15.9.1 DRR Source: Hoosier Energy - Merom Emissions ...... 58

15.9.2 Inventoried SO 2 Sources Included in the Modeling ...... 59 15.10 Modeling Results ...... 59 16.0 - Duke - Cayuga Generating Station (Source ID 18-165-00001) ...... 61 16.1 Source Description ...... 61 16.2 Characterization of Modeled Area ...... 61 iv

16.3 Background Concentrations ...... 62 16.4 Modeling Methodology ...... 62 16.4.1 Model Selection ...... 62 16.4.2 Model Options ...... 62 16.4.3 AERMAP ...... 63 16.5 Meteorological Data ...... 63 16.5.1 AERMET ...... 63 16.5.2 Wind Rose ...... 64 16.5.3 AERMINUTE/AERSURFACE ...... 64 16.6 Receptor Grid and Modeling Domain ...... 65 16.7 Stack Heights ...... 66

16.8 Temporally Varying Seasonal 1-Hour SO 2 Background ...... 66

16.9 SO 2 Emissions Included in the Modeling Analysis ...... 67 16.9.1 DRR Source: Duke - Cayuga Emissions ...... 67

16.9.2 Inventoried SO 2 Sources Included in the Modeling ...... 68 16.10 Modeling Results ...... 68

List of Tables

Table 2.1 - Indiana Sources Subject to the Data Requirements Rule ...... 2 Table 6.1 - National Weather Service Stations/Onsite Meteorological Stations ...... 6 Table 7.1 - Indiana DRR Sources and Nearby Background Monitoring Sites ...... 8

Table 11.1 – SABIC 99th Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) ..... 13 Table 11.2 – SABIC NWS Stations/Onsite Meteorological Stations ...... 15

Table 11.3 – SABIC 99th Percentile Temporally Varying Seasonal SO 2 Background Values (ppb) ...... 18 Table 11.4 – SABIC Modeling Source Inventory ...... 20 Table 11.5 – SABIC Modeling Results ...... 21

Table 12.1 – Lake County 99th Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) ...... 24 Table 12.2 - Lake County Urban Population ...... 25 v

Table 12.3 – Lake County NWS Stations/Onsite Meteorological Stations ...... 26

Table 12.4 – Lake County Hammond Monitor 99th Percentile Temporally Varying Seasonal SO 2 Background Values (ppb) ...... 29

Table 12.5 – Lake County Gary - IITRI 99th Percentiles Temporally Varying Seasonal SO 2 Background Values (ppb) ...... 30 Table 12.6 - Lake County Modeling Inventory ...... 32 Table 12.7 – Lake County Modeling Results ...... 33

Table 13.1 – Duke – Gallagher 99th Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) ...... 35 Table 13.2 – Duke - Gallagher NWS Stations/Onsite Meteorological Stations ...... 37

Table 13.3 – Duke – Gallagher 99th Percentile Temporally Varying Seasonal SO 2 Background Values (ppb) ...... 40 Table 13.4 – Duke – Gallagher Modeling Source Inventory ...... 41 Table 13.5 - Duke – Gallagher Modeling Results...... 42

Table 14.1 – NIPSCO – Schahfer 99th Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) ...... 44 Table 14.2 – NIPSCO – Schahfer NWS Stations/Onsite Meteorological Stations ...... 46

Table 14.3 – NIPSCO – Schahfer 99th Percentile Temporally Varying Seasonal SO 2 Background Values (ppb) ...... 49 Table 14.4 – NIPSCO – Schahfer Modeling Source Inventory ...... 50 Table 14.5 – NIPSCO – Schahfer Modeling Results ...... 50

Table 15.1 – Hoosier Energy – Merom 99th Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) ...... 53 Table 15.2 – Hoosier Energy – Merom NWS Stations/Onsite Meteorological Stations ...... 55

Table 15.3 – Hoosier Energy – Merom 99th Percentile Temporally Varying Seasonal SO 2 Background Values (ppb) ...... 58 Table 15.4 – Hoosier Energy – Merom Modeling Source Inventory ...... 59 Table 15.5 – Hoosier Energy – Merom Modeling Results ...... 60

Table 16.1 – Duke – Cayuga 99th Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) ...... 62 Table 16.2 – Duke – Cayuga NWS Stations/Onsite Meteorological Stations ...... 64

Table 16.3 – Duke – Cayuga 99th Percentile Temporally Varying Seasonal SO 2 Background Values (ppb) ...... 67 Table 16.4 – Duke – Cayuga Modeling Source Inventory ...... 68 vi

Table 16.5 – Duke – Cayuga Modeling Results ...... 68

List of Figures

Figure 11.1 - SABIC Innovative Plastics and Surrounding Area ...... 12 Figure 11.2 – SABIC 3-km Radius to Determine Auer Land Use ...... 14 Figure 11.3 - Evansville 3-year Cumulative Wind Rose (2013 – 2015) ...... 15 Figure 11.4 – SABIC Receptor Grid ...... 17 Figure 11.5 – SABIC Modeling Results ...... 21 Figure 12.1 - Lake County DRR Sources and Surrounding Area ...... 23 Figure 12.2 – Gary-IITRI 3-year Cumulative Wind Rose (2013 – 2015) ...... 26 Figure 12.3 – Lake County Receptor Grid ...... 28 Figure 12.4 – Lake County Modeling Results ...... 33 Figure 13.1 – Duke - Gallagher and Surrounding Area ...... 34 Figure 13.2 – Duke – Gallagher 3-kilometer Radius to Determine Auer Land Use ...... 36 Figure 13.3 - Louisville 3-year Cumulative Wind Rose (2013 – 2015) ...... 37 Figure 13.4 – Duke – Gallagher Receptor Grid ...... 39 Figure 13.5 – Duke - Gallagher Modeling Results ...... 42 Figure 14.1 - NIPSCO - Schahfer and Surrounding Area ...... 43 Figure 14.2 – NIPSCO – Schahfer 3-km Radius to Determine Auer Land Use ...... 45 Figure 14.3 - South Bend 3-year Cumulative Wind Rose (2012 – 2014) ...... 46 Figure 14.4 – NISPCO - Schahfer Receptor Grid ...... 48 Figure 14.5 – NIPSCO - Schahfer Modeling Results ...... 51 Figure 15.1 – Hoosier Energy - Merom and Surrounding Area ...... 52 Figure 15.2 – Hoosier Energy – Merom 3-km Radius to Determine Auer Land Use ...... 54 Figure 15.3 - Evansville 3-year Cumulative Wind Rose (2013 – 2015) ...... 55 Figure 15.4 – Hoosier Energy – Merom Receptor Grid...... 57 Figure 15.5 – Hoosier Energy - Merom Modeling Results ...... 60 Figure 16.1 - Duke - Cayuga and Surrounding Area ...... 61 Figure 16.2 – Duke – Cayuga 3-km Radius to Determine Auer Land Use ...... 63 Figure 16.3 – Indianapolis 3-year Cumulative Wind Rose (2012 – 2014) ...... 64 vii

Figure 16.4 – Duke – Cayuga Receptor Grid ...... 66 Figure 16.5 – Duke - Cayuga Modeling Results ...... 69

List of Enclosures

Enclosure 1: 1-Hour Sulfur Dioxide Background Determination

Enclosure 2: Lake County DRR Source Modeling Inventory Enclosure 3: Carmeuse Commissioner’s Order #2016-04 Enclosure 4: SABIC Commissioner’s Order #2016-03

viii

MODELING TECHNICAL SUPPORT DOCUMENT FOR PRELIMINARY DESIGNATION RECOMMENDATIONS

DATA REQUIREMENTS RULE (ROUND 3) FOR THE

2010 PRIMARY 1-HOUR SULFUR DIOXIDE (SO 2) NATIONAL AMBIENT AIR QUALITY STANDARD

1.0 1-Hour SO2 NAAQS and Designation Process

The United States Environmental Protection Agency (U.S. EPA) established the 1-hour primary sulfur dioxide (SO 2) National Ambient Air Quality Standard (NAAQS) of 75 parts per billion (ppb) as published in the Federal Register (FR) on June 22, 2010 (75 FR 35519). This standard is based on the 3-year average of the annual 99 th percentile of the 1-hour daily maximum concentrations. For air quality modeling purposes, the Indiana Department of Environmental

Management (IDEM), Office of Air Quality (OAQ) uses an equivalent 1-hour SO 2 NAAQS of 196.2 micrograms per cubic meter (µg/m 3) as stated in 76 FR 69051. This is based on the 3-year th average of the annual 99 percentile of the 1-hour daily maximum modeled SO 2 concentrations, representing the fourth high of the 1-hour daily maximum SO 2 modeled concentrations.

Implementation of the standard began in 2013, when U.S. EPA made initial designations based on 2010-2012 monitoring data (78 FR 47191). Subsequently, on March 2, 2015, U.S. EPA entered into a consent decree with the Sierra Club and the Natural Resources Defense Council establishing a timeline for the completion of air quality characterizations designations in all remaining areas of the country. The court order directed U.S. EPA to complete the designations in three additional rounds: Round 2 by July 2, 2016 (81 FR 45039), Round 3 by December 31, 2017, and Round 4 by December 31, 2020.

Round 3 and 4 designations are implemented through U.S. EPA’s SO 2 Data Requirements Rule (DRR) (80 FR 51051). Round 3 designations apply to source areas that opt to characterize SO 2 through modeling and have not implemented ambient air monitoring by January 1, 2017. Round

4 designations apply to source areas that opt to characterize SO 2 by having implemented new ambient air monitoring by January 1, 2017. In addition, sources may opt to take permanent

federally enforceable emission limits in order to reduce SO 2 emissions to below the DRR threshold of 2,000 tons per year.

Page 1 of 69

2.0 Data Requirements Rule

As stated above, Round 3 designations are implemented through U.S. EPA’s SO 2 DRR. Under this rule, SO 2 should be characterized in the vicinity of sources that had actual emissions in 2014 of 2,000 tons or more, or have been identified by IDEM or U.S. EPA “as requiring further air quality characterization.”

Requirements specific to the DRR were followed in order to implement the 1-hour SO 2 NAAQS. Indiana identified 11 sources within the state that met the criteria established in the DRR. This list of sources was submitted to U.S. EPA – Region V on January 7, 2016. On March 25, 2016, U.S. EPA subsequently identified six additional sources meeting the criteria for air quality characterizations under the DRR. Five of these sources were “consent decree” sources and were designated unclassifiable/attainment under Round 2 (81 FR 45039). The sixth source, U.S. Mineral Products (U.S. Minerals) was listed by U.S. EPA as subject to the DRR due to concern for air quality in the area. All DRR sources, the counties they reside and their 2014 SO 2 emissions are listed in Table 2.1.

Table 2.1 - Indiana Sources Subject to the Data Requirements Rule

Facility County 2014 SO 2 Emissions (tons) Duke – Gallagher Floyd 3,524 Duke – Gibson Gibson Round 2 Source a U.S. Mineral Products (Isolatek) Huntington < 2,000 b NIPSCO – R.M. Schahfer Jasper 8,412 IKEC–Clifty Creek Generating Station Jefferson Round 2 Source a ArcelorMittal – USA Lake 2,163 Coke Energy Lake 4,952 U.S. Steel – Gary Works Lake 3,285 NIPSCO - Michigan City LaPorte Round 2 Source a ArcelorMittal - Burns Harbor Porter 12,189 SABIC Innovative Plastics Posey 4,030 Vectren—A.B. Brown Generating Station Posey Round 2 Source a AEP - Rockport Spencer Round 2 Source a Hoosier Energy – Merom Sullivan 3,318 Duke – Cayuga Vermillion 3,448 Alcoa Warrick Power Plant Warrick 4,993 Alcoa Warrick Operations Plant Warrick 3,500 c a IDEM completed characterization for this source under Round 2 designation requirements. U.S. EPA issued final Round 2 designations on June 30, 2016 (81 FR 45039). b Added by U.S. EPA. c Alcoa Warrick Operations shut down its smelter operations on March 31, 2016, reducing SO 2 emissions to < 1 ton source-wide. Page 2 of 69

As per the requirements of the DRR, air agencies were required to indicate whether they will rely on 1) air quality modeling, 2) ambient monitoring or 3) establishing a limit of a source’s total

SO 2 emissions to below 2,000 tons per year, to characterize air quality in the area surrounding the DRR sources. Indiana reviewed each source and determined that eight sources will conduct air dispersion modeling to characterize air quality including, where appropriate, modeling non- DRR sources. One source, ArcelorMittal – Burns Harbor, opted to rely on ambient monitoring to characterize air quality (see Section 10.0 and transmittal Attachment 5). For U.S. Mineral Products (Isolatek), Indiana disagrees with U.S. EPA on its inclusion as being subject to the DRR (see transmittal Attachment 3). Lastly, for Alcoa Warrick Operations and Alcoa Warrick Power Indiana feels that these two facilities and the surrounding area should be designated attainment based on historical SO 2 ambient monitoring showing attainment of the SO 2 standard and the fact that the Operations Plant shut down their aluminum smelting operations on March

31, 2016 and has negligible SO 2 emissions as a result of the shutdown (see transmittal Attachment 4).

U.S. EPA has established deadlines for each step of the 1-hour SO 2 designation process in the DRR. Indiana met the first deadline by submitting its list of DRR sources on January 7, 2016.

• January 15, 2016 - States were required to submit their list of SO 2 sources for characterizing air quality under the DRR to U.S. EPA. • July 1, 2016 – States were required to submit modeling protocols for sources characterizing air quality in the area with air dispersion modeling. • July 1, 2016 – States were required to submit Annual Monitoring Network Plans that detailed modifications to SO 2 monitors intended to satisfy the DRR. • January 1, 2017 – SO 2 monitors intended to satisfy the DRR are required to be operational. • January 13, 2017 – States electing to characterize air quality by air dispersion modeling are required to provide modeling analyses to U.S. EPA. • January 13, 2017 – Federally enforceable and permanent emission limits to keep source emissions below 2,000 tons of SO 2 must be adopted and effective. • August 2017 – Expected date by which U.S. EPA will notify states of intended designations. • December 2017 – Date by which U.S. EPA will complete final designations for the majority of the country. • August 2019 – Approximate due date for state attainment plans for areas designated nonattainment in 2017. • May 2020 – Required certification of 2019 monitoring data; states have the opportunity to provide updated state recommendations to U.S. EPA. • August 2020 – Expected date by which U.S. EPA would notify states of intended designations for reminder of the country not yet designated.

Page 3 of 69

• December 2020 – Date by which the U.S. EPA would complete final designations for the remainder of the country. • August 2022 – Approximate due date for state attainment plans for areas designated nonattainment in 2020.

3.0 Methodology for DRR Air Quality Modeling

The modeling methodology resembles modeling used to evaluate New Source Review (NSR) and Prevention of Significant Deterioration (PSD) sources. However, U.S. EPA provided further guidance in order to conduct an appropriate air dispersion modeling analysis to support 1-hour

SO 2 designation recommendations. U.S. EPA’s SO 2 NAAQS Designations Modeling Technical Assistance Document (TAD) guidance has several recommendations for modeling methodology for determining attainment designations, including:

1) Use of actual emissions to assess modeled concentrations to reflect current air quality.

2) Use of three years of modeling results to calculate a simulated 1-hour SO 2 design value consistent with the 3-year monitoring period to develop 1-hour SO 2 design values. 3) Placement of receptors only in locations where an air quality monitor could be placed.

• Based on the SO 2 NAAQS Designations Modeling TAD, Section 4.2; Indiana placed modeling receptors only where feasible to place a monitor. Therefore, in bodies of water or an area where monitor siting criteria would not be reasonably met, Indiana did not place receptors. • Indiana matched up the modeling domain with Google map projections to ensure the proximity of the receptors to shorelines and have provided receptor/mapping details for each modeling analysis. 4) Use of actual stack heights rather than relying on Good Engineering Practice (GEP) stack heights when modeling actual emissions.

Indiana followed U.S. EPA’s designation modeling recommendations to conduct 1-hour SO 2 modeling to determine whether there are modeled violations of the 1-hour SO 2 NAAQS. th Modeling results looked at the 4 high maximum daily 1-hour SO 2 concentrations averaged over the 3-year modeled period with representative temporally varying seasonal SO 2 background concentrations included within the AERMOD modeling run to determine the attainment status of the area in the vicinity of the DRR source.

Page 4 of 69

4.0 Model Selection for DRR Modeling

4.1 AERMOD Dispersion Model In accordance with Appendix A of Appendix W to 40 Code of Federal Regulations (CFR) Part 51, Indiana used the American Meteorological Society/Environmental Protection Agency

Regulatory Model (AERMOD) version 15181 for all dispersion modeling. U.S. EPA’s SO 2 NAAQS Designations Modeling TAD, specific to attainment designation modeling, recommended using actual stack heights when modeling actual emissions instead of following the GEP stack height requirement. BPIPPRIME was used to account for any building downwash concerns.

4.2 AERMAP The AERMOD terrain preprocessor mapping program, AERMAP, was used to determine all the terrain elevation heights for each receptor, building, and source locations using the Universal Transverse Mercator (UTM) coordinate system. The most recent AERMAP version 11103 assigned the elevations from the National Elevation Dataset (NED) using the North American Datum (NAD) 1983 as recommended in 40 CFR Part 51, Appendix W, Revision to the Guideline on Air Quality Models and later revised in the “AERMOD Implementation Guide.”

4.3 Land Use Determination The Auer Land Use Classification Scheme was used to determine land use in the area of each source, pursuant to 40 CFR Part 51, Appendix W section 7.2.3(c). Land use types were classified within a 3 kilometer radius about the source. If land use types I1 (heavy industrial), I2 (light moderate industrial), C1 (commercial), R2-R3 (compact residential) account for over 50 percent of the total land area, urban dispersion coefficients were used. If not, the rural dispersion coefficients were used.

5.0 Receptor Grid and Modeling Domain

The receptor grid and modeling domain was based on guidance provided in the memorandum “Updated Guidance for Area Designations for the 2010 Primary Sulfur Dioxide National

Ambient Air Quality Standards”, dated March 20, 2015 and the SO 2 NAAQS Designations Modeling TAD. Indiana used a multi-nested rectangular receptor grid with appropriate spacing of receptors based on the distance from the modeled emission points to detect significant concentration gradients. Indiana has conducted exploratory modeling on each of the DRR

sources and did not find maximum modeled 1-hour SO 2 impacts or DRR source-culpable modeled violations that extended out beyond 10 kilometers. In situations where multiple sources covered by the DRR were evaluated in the same area, the modeling domain extended to include

Page 5 of 69

all sources and the appropriate distances to model maximum 1-hour SO 2 impacts to determine attainment designations for the area. Indiana generally used the following multi-nested rectangular receptor grid in all cases, with additional receptors added as needed:

• Receptor spacing at the fence line for each facility placed every 50 meters. • Receptor spacing at 100 meters out to a distance of 3,000 meters (3 kilometers) beyond each facility (grid was extended if modeling results warranted). • Receptor spacing at 250 meters out to a distance of 5,000 meters (5 kilometers) beyond each facility (grid was extended if modeling results warranted). • Receptor spacing at 500 meters out to a distance of 10,000 meters (10 kilometers) beyond each facility (grid was extended if modeling results warranted).

6.0 Meteorology

6.1 AERMET

As stated in 40 CFR Part 51, Appendix W, section 8.3.1.2 and the SO 2 NAAQS Designations Modeling TAD, Indiana used three years (2012-2014 or 2013-2015) of National Weather Service (NWS) and on-site surface data and upper air meteorological data processed with the latest version of the AERMOD meteorological data preprocessor program AERMET (version 15181). Table 6.1 below lists the modeled facilities as mentioned in the DRR and the corresponding surface and upper air meteorological stations used to conduct modeling.

Table 6.1 - National Weather Service Stations/Onsite Meteorological Stations DRR Facility Surface Meteorology Upper Air Meteorology SABIC Innovative Plastics Evansville, IN NWS Lincoln, IL NWS Hoosier Energy - Merom Duke – Gallagher Louisville, KY NWS Wilmington, OH NWS Arcelormittal – USA Gary-IITRI onsite Coke Energy meteorological data Lincoln, IL NWS U.S. Steel – Gary Works processed with South ArcelorMittal Burns Harbor Bend, IN NWS NIPSCO – R.M. Schahfer South Bend, IN NWS Lincoln, IL NWS Duke –Cayuga Indianapolis, IN NWS Lincoln, IL NWS

Indiana requested on November 9, 2016 for concurrence by U.S. EPA for the use of the adjusted surface friction velocity (ADJ_U*) Beta option in order to more accurately model 1-hour SO 2 concentrations from DRR sources located in Lake County. On December 20, 2016, U.S. EPA finalized “Revisions to the Guidelines on Air Quality Models, Enhancements to the AERMOD Dispersion Modeling System and Incorporation of Approaches to Address Ozone and Fine Page 6 of 69

Particulate Matter”. This rule approved ADJ_U* as a regulatory option and was used in the DRR modeling for Lake County.

6.2 AERMINUTE/AERSURFACE The 1-minute wind speeds and wind directions, taken from the Automated Surface Observing System (ASOS) NWS stations and onsite meteorological stations, were processed with the U.S. EPA’s 1-minute data processor AERMINUTE (version 15272) program.

The U.S. EPA’s AERSURFACE (version 13016) program was used to determine the surface characteristics; albedo, Bowen ratio, and surface roughness for the Indianapolis, Evansville, South Bend, Indiana and Louisville, Kentucky NWS meteorological tower locations. Surface characteristics were determined at each NWS location for each of 12 wind direction sectors with a recommended default radius of one kilometer.

The albedo and the Bowen ratio surface characteristics were adjusted during the three winter months of January, February, and December in accordance with the U.S. EPA Region V document, “Regional Meteorological Data Processing Protocol,” dated May 6, 2011. Additionally, a dry or wet Bowen ratio value was used during months when soil moisture conditions were abnormally dry or wet; otherwise the Bowen ratio value for average soil moisture conditions was used. The surface roughness value for snow cover was used if more than half of the month had days with at least one inch of snow on the ground. Otherwise, the no snow cover surface roughness value was used.

7.0 SO 2 Background Concentrations

The modeling of all DRR sources used adjusted temporally varying seasonal background concentrations or concentrations without upwind major source SO 2 impacts. Each source used 1- hour SO 2 monitoring data, taken from nearby monitors, considered representative of background concentrations for the area. Since most SO 2 monitoring sites located in the state are downwind of large SO 2 sources, impacts from the upwind direction of the large SO 2 source were removed from the monitoring data since those sources were included in the modeling inventory. The 99 th percentile SO 2 concentrations by season (winter, spring, summer and fall) for each hour of the day were calculated to determine the temporally varying seasonal SO 2 background, which were directly input into the model and were part of the final modeled results. This procedure was used to prevent double counting of SO 2 sources within the background concentration values used for this attainment designation modeling.

Temporally varying seasonal SO 2 background concentrations were developed in accordance with the recommended U.S. EPA guidance for establishment of such background concentrations in Page 7 of 69

Section 8.2 of 40 CFR Part 51, Appendix W and considered appropriate and representative of the area. The latest three years of SO 2 air quality monitoring data (2012-2014 or 2013-2015) were used to develop background concentrations for each of the areas mentioned in the DRR. The procedures used to develop the SO 2 background concentrations are included as Enclosure 1. Table 7.1 shows the DRR facility and corresponding 1-hour SO 2 monitoring sites used for representative background concentrations in the air quality characterization.

Table 7.1 - Indiana DRR Sources and Nearby Background Monitoring Sites Facility County Monitoring Sites SABIC Innovative Plastics Posey Evansville – Buena Vista Duke – Gallagher Floyd New Albany – Green Valley NIPSCO – R.M. Schahfer Jasper Wheatfield – Center St. Hoosier Energy – Merom Sullivan Terre Haute – North Lafayette Road Duke – Cayuga Vermillion Fountain County -North of State Road 234 ArcelorMittal – USA Coke Energy Lake Gary-IITRI and Hammond U.S. Steel – Gary Works ArcelorMittal - Burns Harbor Porter Dunes Acres Substation

8.0 SO 2 Emissions Sources to be Modeled

8.1 DRR Sources Indiana modeled the hourly continuous emissions monitoring (CEM) data from sources subject to the DRR, where available. Along with the hourly SO 2 emission data, hourly variable stack gas flow rate and temperature of the exhaust stream were modeled, if available. This variation in parameters may influence dispersion characteristics of the exhaust stream and impact modeled 1-

hour SO 2 concentrations.

For those emission sources without continuous emissions data, actual short-term emissions taken from the source’s latest available emissions reporting were used. The SO 2 NAAQS Designations Modeling TAD, Section 5 was referenced to best characterize any temporal and/or seasonal variability of emissions. This would include any seasonal, monthly, or daily variations that can be quantified. Specific emissions characterizations that were modeled will be addressed for each DRR source later in this document.

There are instances where sources emitted less than 2,000 tons of SO 2 in 2014 and are not listed as a DRR source, but are located in the vicinity of a DRR source within the modeling receptor grid. This was considered a cluster source and the source was evaluated along with the DRR

Page 8 of 69

source in the air quality modeling analysis to determine the air quality characterization in the area.

8.2 Inventory Sources Based on the U.S. EPA memo “Additional Clarification Regarding Application of Appendix W

Modeling Guidance for the 1-hour NO 2 National Ambient Air Quality Standard”, dated March 1, 2011, page 16, Indiana is focused on the characterization of air quality within 10 kilometers for

each of the DRR sources. U.S. EPA’s SO 2 NAAQS Designations Modeling TAD Section 4.1, page 7, mentions the number of sources to be explicitly modeled should cause a significant concentration gradient and the number of those sources to be modeled would generally be small.

Indiana developed a list of SO 2 emission sources in the county of the DRR source, as well as larger SO 2 emission sources in adjacent counties and states, as requested by U.S. EPA – Region 5, that were explicitly modeled.

Emission sources near the DRR source were evaluated to determine if those sources could cause

or contribute to a 1-hour SO 2 NAAQS violation. Indiana used the following threshold as a screening method to narrow the focus of sources that could potentially have an impact on designations: sources with SO 2 emissions greater than 250 tons per year and located within 30 kilometers of the DRR source. While this method was applied on an area-by-area basis, Indiana felt this was an accurate representation of air quality in the area, especially since the hourly

seasonal background concentrations adequately captures SO 2 impacts from surrounding sources. IDEM also identified sources with emissions less than 250 tons that were included in DRR modeling due to their proximity within the DRR source receptor grid used in the dispersion modeling. Actual emissions taken from the latest available emissions inventories were modeled for sources identified by these threshold levels to determine air quality characteristics in the area.

8.3 Intermittent Sources Emergency generators, fire pumps, and startup/shutdown emissions were handled consistent to the March 1, 2011 guidance “Additional Clarification Regarding Application of Appendix W

Modeling Guidance for the 1-hour NO 2 NAAQS”, dated March 1, 2011. U.S. EPA recommended using appropriate data based on emissions scenarios that are continuous enough or frequent enough to contribute significantly to the annual distribution of maximum daily 1-hour concentrations. Review of the hours of operations for combustion turbines, emergency generators, startup/shutdown, fire pumps, and other auxiliary operations associated with the sources addressed by the DRR have been determined to operate much less than 500 hours per year and have random and infrequent schedules that cannot be controlled. Indiana feels that the intent of the DRR is to determine the attainment status of the area surrounding large SO 2 emission sources based on actual emissions coming from the large units. As such, this is

Page 9 of 69

Indiana’s main focus of the designation determinations. This approach is consistent with previous 1-hour SO 2 nonattainment and designation modeling submitted by IDEM to U.S. EPA.

9.0 Analysis of Modeling Results

The purpose of this modeling demonstration is to characterize air quality and determine area designations as it relates to attainment of the 1-hour SO 2 NAAQS in accordance with the DRR. The 99 th percentile of the 1-hour daily maximum modeled concentrations represents the fourth

high of the 1-hour daily maximum SO 2 modeled concentrations and are averaged across three 3 years to compare resulting concentrations to the 1-hour SO 2 NAAQS of 75 ppb (196.2 µg/m ).

Modeled concentrations include representative temporally varying seasonal 1-hour SO 2 background values to determine the overall impact from the DRR sources. This resulting

concentration is compared to the 1-hour SO 2 standard to indicate whether a modeled violation of the SO 2 NAAQS has occurred. All concentrations that fall at or below the 1-hour SO 2 NAAQS are determined to attain the standard and the area surrounding the DRR source is recommended as attainment.

10.0 ArcelorMittal – Burns Harbor (Source ID 18-127-00001)

10.1 Source Description ArcelorMittal - Burns Harbor, LLC (Burns Harbor) is a stationary steel works plant for the production of coke, limited coal chemical, molten iron, molten steel, steel slabs, hot rolled steel, steel coils, steel plates, cold rolled and/or coated steel sheet and plate. Specific emission units associated with Burns Harbor include a coke oven process plant, coke by-products recovery plant, blast furnace granulated coal injection system, continuous sintering process plant, two blast furnaces, basic oxygen furnaces shop, slab/plate mill complex, hot strip mill, cold sheet mill operations, power station, service shop and technical maintenance operations and fugitive dust emission operations including sinter plant and blast furnace operations.

10.2 Characterization of Modeled Area Burns Harbor opted to select the monitoring option for air quality characterization in the vicinity of its facility. Therefore, a modeling analysis was conducted to determine the location of maximum modeled 1-hour SO 2 impacts near the facility. Once the location of maximum impacts was determined, Burns Harbor located an ambient air monitor near that location in order to accurately measure the SO 2 impacts from Burns Harbor and nearby SO 2 sources to compare with the 1-hour SO 2 NAAQS.

Page 10 of 69

10.3 Summary of DRR Monitoring Approach

Burns Harbor and IDEM completed a modeling analysis and SO 2 Monitor Quality Assurance and Project Plan (QAPP) to site an SO 2 monitor and submitted both to U.S. EPA- Region 5 on June 10, 2016. On August 5, 2016, U. S. EPA approved the analysis and general monitor site location based on “hot spot” modeling to determine the maximum modeled 1-hour SO 2 concentration. Burns Harbor procured monitoring equipment and obtained, from the Port of Indiana, a lease for land. U.S. EPA approved IDEM’s monitoring network for 2017 on October

31, 2016, which included the Burns Harbor SO 2 monitor. Burns Harbor was able to construct a concrete pad and shelter, set up and calibrate the equipment in early December 2016 and began operation of the monitor in mid-December, well ahead of the January 1, 2017 deadline. Clean Air Engineering completed testing of the communications system and verified calibration of all monitoring equipment. This monitor has been assigned AQS Identification number: 18-127- 0028 and was operational on or before January 1, 2017. The monitoring network, consisting of the ArcelorMittal – Burns Harbor and the Dunes Acres Substation (AQS ID #18-127-0011) monitors meets the DRR requirement.

11.0 SABIC Innovative Plastics (Source ID 18-129-00002)

11.1 Source Description SABIC Innovative Plastics Mt. Vernon, LLC (SABIC) is a plastics manufacturing facility. SABIC produces plastics for industries such as automotive, consumer electronics and medical devices.

SABIC is retrofitting their facility with a cogeneration (CoGen) plant that will use natural gas to create a majority of the steam for the site. Currently, SABIC’s coal-fired boilers provide approximately 40 percent of the facility’s steam. The U.S. EPA recently issued a new Maximum Achievable Control Technology (MACT) standard for industrial, commercial, and institutional boilers. SABIC is building their CoGen plant to address those standards. Significant SO 2 emission reductions are a byproduct of this project as several coal-fired boilers at SABIC were shut down once the project became fully operational by the end of December of 2016.

SABIC was identified as a Data Requirements Rule (DRR) source based on their actual 2014

SO 2 emissions of 4,030 tons exceeding the DRR threshold of 2,000 tons of SO 2. While the CoGen project helped SABIC realize significant SO 2 emission reductions, potential SO 2 emissions from the facility were still above 2,000 tons. The modeling option was chosen to address the DRR.

Initial modeling, using actual emissions data from 2014, showed higher modeled 1-hour SO 2 concentrations. However, after discussions with SABIC, it was decided they would request a Page 11 of 69

Commissioner’s Order to establish plant-wide SO 2 emission limits that would be federally enforceable and permanent and would model attainment of the 1-hour SO 2 standard.

11.2 Characterization of Modeled Area SABIC is located at 1 Lexan Lane, Mt. Vernon, Indiana, less than a mile from the Ohio River in Black Township, Posey County, Indiana. A map of the area surrounding the SABIC facility and the township in which SABIC is located is shown below in Figure 11.1.

Figure 11.1 - SABIC Innovative Plastics and Surrounding Area

11.3 Background Concentrations

The nearest 1-hour SO 2 monitored concentrations were taken from the Evansville – Buena Vista monitor (AQS #18-163-0021). The 99 th percentile values from 2013 through 2015 and the 3- year design value are listed below in Table 11.1. Concentrations are well below the 1-hour SO 2 standard.

Page 12 of 69

th Table 11.1 – SABIC 99 Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) Monitoring Site 2013 2014 2015 2013-2015 Evansville – Buena Vista 18.6 32.3 18 23

11.4 Modeling Methodology The SABIC DRR modeling methodology resembles modeling used to evaluate New Source Review (NSR) and Prevention of Significant Deterioration (PSD) sources. However, Indiana has relied on U.S. EPA guidance “EPA’s SO 2 NAAQS Designations Modeling Technical Assistance Document” in order to conduct an appropriate air dispersion modeling analysis for SABIC to support 1-hour SO 2 designation recommendations.

11.4.1 Model Selection In accordance with Appendix A of Appendix W to 40 Code of Federal Regulations (CFR) Part 51, Indiana used the American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD) version 15181. BPIPPRIME was used to account for any building downwash concerns.

11.4.2 Model Options All regulatory default options within AERMOD were used to determine the air quality characteristics surrounding SABIC. The area is considered primarily rural, based on the Auer’s Classification Land Use methodology with a vast majority of the land use types within 3 kilometers of SABIC, classified as metropolitan natural (A1), agricultural rural (A2), water surfaces (A5) and estate residential (R4). Therefore, a rural classification was used, as provided for in the Guideline on Air Quality Models, Section 7.2.3 (EPA, 2005b). No variation of the population selection was necessary. Figure 11.2 shows the 3-kilometer radius area surrounding SABIC that was analyzed to determine the land use classification.

Page 13 of 69

Figure 11.2 – SABIC 3-km Radius to Determine Auer Land Use

11.4.3 AERMAP The AERMOD terrain preprocessor mapping program, AERMAP, was used to determine all the terrain elevation heights for each receptor, building, and source locations using the Universal Transverse Mercator (UTM) coordinate system. The most recent AERMAP version 11103 assigned the elevations from the National Elevation Dataset (NED) using the North American Datum (NAD) 1983 as recommended in the, “40 CFR Part 51, Revision to the Guideline on Air Quality Models” Appendix W and later revised in the “AERMOD Implementation Guide.”

11.5 Meteorological Data 11.5.1 AERMET

As stated in 40 CFR Part 51, Appendix W, section 8.3.1.2 and the SO 2 NAAQS Designations Modeling TAD, Indiana used 2013-2015 National Weather Service (NWS) surface and upper air meteorological data processed with the latest version of the AERMOD meteorological data

Page 14 of 69

preprocessor program AERMET (version 15181). Table 11.2 below lists surface and upper air meteorological stations used to conduct modeling.

Table 11.2 – SABIC NWS Stations/Onsite Meteorological Stations Facility Surface Meteorology Upper Air Meteorology SABIC Innovative Plastics Evansville, IN NWS Lincoln, IL NWS

11.5.2 Wind Rose The Evansville National Weather Service (NWS) surface meteorological data and the Lincoln, Illinois upper air meteorological data taken from 2013 through 2015 was used to determine the meteorological conditions for the area surrounding SABIC in AERMOD. The Evansville NWS wind rose for the 3-year modeled period 2013-2015 is shown as Figure 11.3 below. The Evansville NWS wind rose depicts the predominant wind direction as from the southwest for the 3-year modeled period of 2013-2015.

Figure 11.3 - Evansville 3-year Cumulative Wind Rose (2013 – 2015)

11.5.3 AERMINUTE/AERSURFACE The 1-minute wind speeds and wind directions, taken from the Automated Surface Observing System (ASOS) NWS stations and onsite meteorological stations, were processed with the U.S. EPA 1-minute data processor program AERMINUTE version 15272.

Page 15 of 69

The U.S. EPA program AERSURFACE version 13016 was used to determine the surface characteristics; albedo, Bowen ratio, and surface roughness for the Evansville, Indiana NWS meteorological tower location. Surface characteristics were determined at the NWS location for each of 12 wind direction sectors with a recommended default radius of one kilometer. The albedo and the Bowen ratio surface characteristics were adjusted during the three winter months of January, February, and December in accordance with the U.S. EPA Region V document, “Regional Meteorological Data Processing Protocol,” dated May 6, 2011. Additionally, a dry or wet Bowen ratio value was used during months when soil moisture conditions were abnormally dry or wet; otherwise the Bowen ratio value for average soil moisture conditions was used. The surface roughness value for snow cover was used if more than half of the month had days with at least one inch of snow on the ground. Otherwise, the no snow cover surface roughness value was used.

11.6 Receptor Grid and Modeling Domain The receptor grid and modeling domain was based on guidance provided in the memorandum “Updated Guidance for Area Designations for the 2010 Primary Sulfur Dioxide National

Ambient Air Quality Standards”, dated March 20, 2015 and the SO 2 NAAQS Designations Modeling TAD. Indiana used a multi-nested rectangular receptor grid with appropriate spacing of receptors based on the distance from the modeled emission points to detect significant concentration gradients. The modeling domain extended out to include all sources and the

appropriate distances to model maximum 1-hour SO 2 impacts to determine attainment designations for the area. Indiana used the following multi-nested rectangular receptor grid which are listed below and depicted in Figure 11.4:

• Receptor spacing at the fence line for the DRR facility was placed every 50 meters. • Receptor spacing at 100 meters was placed out to a distance of 3,000 meters (3 kilometers) beyond the DRR facility. • Receptor spacing at 250 meters was placed out to a distance of 5,000 meters (5 kilometers) beyond the DRR facility. • Receptor spacing at 500 meters was placed out to a distance of 10,000 meters (10 kilometers) beyond the DRR facility.

Page 16 of 69

Figure 11.4 – SABIC Receptor Grid

The SABIC property is fully fenced and has regular security patrols to keep unauthorized people off the property. Since this is the case, receptors were placed along the property lines.

11.7 Stack Heights The use of actual stack heights rather than relying on Good Engineering Practice (GEP) stack heights when modeling actual emissions was utilized in the analysis per the SO 2 NAAQS Designations Modeling TAD.

11.8 Temporally Varying Seasonal 1-Hour SO 2 Background

Temporally varying seasonal SO 2 background concentrations were developed in accordance with the recommended U.S. EPA guidance for establishment of such background concentrations in Section 8.2 of 40 CFR Part 51, Appendix W and considered appropriate and representative of the area. The latest three years of SO 2 air quality monitoring data (2013-2015) was used.

Page 17 of 69

th The 99 percentile SO 2 concentrations by season (winter, spring, summer and fall) for each hour of the day were calculated to determine the temporally varying seasonal SO 2 background, which were directly input into the model and were part of the final modeled results.

Temporally varying seasonal 1-hour SO 2 background concentrations were taken from the Evansville – Buena Vista Road monitor for 2013 - 2015. The hourly seasonal SO 2 values used for representative background concentrations for the area surrounding SABIC are listed below in Table 11.3.

Table 11.3 – SABIC 99 th Percentile Temporally Varying

Seasonal SO 2 Background Values (ppb) Hr 1 Hr 2 Hr 3 Hr 4 Hr 5 Hr 6 Hr 7 Hr 8

Winter 6.30 4.83 4.63 4.36 5.77 4.84 4.70 7.39 Spring 5.12 3.89 4.09 3.98 3.40 4.20 6.83 7.59 Summer 2.70 2.48 1.00 1.00 1.96 2.65 2.80 5.55 Fall 4.44 4.52 4.50 4.50 4.80 4.60 4.97 5.70

Hr 9 Hr 10 Hr 11 Hr 12 Hr 13 Hr 14 Hr 15 Hr 16

Winter 9.29 10.42 9.20 10.67 11.55 17.57 8.71 16.01 Spring 9.99 9.84 11.89 11.65 7.94 9.89 8.39 8.55 Summer 9.93 11.05 8.50 9.02 7.34 5.65 5.49 5.16 Fall 7.55 10.68 11.37 11.21 10.39 12.92 9.11 7.56

Hr 17 Hr 18 Hr 19 Hr 20 Hr 21 Hr 22 Hr 23 Hr 24 Winter 9.94 16.85 8.28 6.67 5.74 6.58 6.79 7.98 Spring 11.04 12.53 9.99 8.40 5.81 3.92 7.04 6.65 Summer 4.11 6.99 5.88 4.05 3.36 2.45 3.58 2.19 Fall 8.20 6.95 5.23 8.60 5.70 4.68 4.46 4.40

11.9 SO 2 Emissions Included in the Modeling Analysis 11.9.1 DRR Source: SABIC Emissions

As a result of the CoGen project, a number of SO 2 emission units will shut down. The unit that will still have significant SO 2 emissions is the COS Vent Oxidizer. SABIC has 16 carbon monoxide (CO) reactors, or generators, that are used to manufacture carbon monoxide. The CO generators are located in the phosgene process area. CO is generated by combusting coke (a petroleum-based material that consists mostly of carbon, with minor amounts of sulfur as an impurity) in the CO generators under low-oxygen conditions. Because the coke contains low levels of sulfur, the raw CO from the CO generators contains sulfur-containing impurities

Page 18 of 69

(carbonyl sulfide, carbon disulfide, and hydrogen sulfide). These impurities need to be removed prior to the next step in the manufacturing process, where CO is combined with chlorine to make phosgene.

The raw CO is purified by passing it through one of several carbon adsorbers. At the outlet of the adsorber, a gas chromatograph measures the concentrations of the sulfur-containing compounds in the purified CO. Once a certain level of sulfur-containing compounds is detected, the flow of raw CO is switched to another adsorber and the spent adsorber is regenerated by desorbing the sulfur-containing compounds with heated nitrogen.

The adsorbed regeneration gas (primarily nitrogen, with low levels of sulfur-containing compounds) is then vented to either the COS Vent Oxidizer or the COS Flare. The regeneration gas passes through a valve that directs the flow to either the COS Vent Oxidizer or the COS Flare, but cannot direct the flow to both simultaneously. The COS Vent Oxidizer is the primary control device; the COS Flare serves as a back-up to the COS Vent Oxidizer or during safety interlock of the system. Both the COS Vent Oxidizer and COS Flare eliminate the sulfur- containing compounds in the regeneration gas by thermal combustion.

Since SO 2 emissions can be routed to either the COS Vent Oxidizer or COS Flare, modeling was performed for both scenarios to determine the worst-case dispersion. Other ancillary sources such as the liquid waste boilers were included in the inventory. Most of the other ancillary sources have small SO 2 emissions (i.e. generators and fire pumps) but were included in the modeling. All SABIC emission limits were based on fuel usage and emissions calculations taken from U.S. EPA’s AP-42 emission factors. All the emission limits that are in the Commissioner’s Order #2016-03 have been represented in the modeling analysis. The Commissioner’s Order can be found in Enclosure 4.

11.9.2 Inventoried SO 2 Sources Included in the Modeling

SO 2 sources from the surrounding area were evaluated to determine if their SO 2 emissions had a potential impact on the air quality surrounding SABIC, beyond what is captured through background monitoring data. The latest available actual emissions were input for some of the inventory sources.

CountryMark had a reduction in SO 2 emissions as a result of installing equipment to recover the vacuum off-gas (a refinery fuel gas) rather than combusting it in the crude heater. The recovered vacuum off-gas is routed to the refinery amine unit and sulfur recovery unit where a high percentage of the sulfur compounds are converted to molten sulfur. Since this was the case, the

2015 emissions were used in the modeling analysis. A.B. Brown was modeled with the SO 2 emission limits listed in their Commissioner’s Order #2016-01. Midwest Fertilizer is still under

Page 19 of 69

construction and is not in full operation so an SO 2 emission rate taken from their permit was modeled. Table 11.4 lists the sources that were included in the AERMOD run to determine overall air quality characteristics.

Table 11.4 – SABIC Modeling Source Inventory

Source Source ID Location SO 2 Emissions (tpy) CountryMark 129-00037 Posey County 65.7 A.B. Brown 129-00010 Posey County Emission Limits a Midwest Fertilizer 129-00059 Posey County 1.3 a A.B. Brown established SO 2 emission limits in response to Round 2 designation requirements

11.10 Modeling Results The 99 th percentile of the 1-hour daily maximum modeled concentrations represents the fourth

high of the 1-hour daily maximum SO 2 modeled concentrations and were averaged across three 3 years to compare resulting concentrations to the 1-hour SO 2 NAAQS of 75 ppb (196.2 µg/m ). Modeled concentrations include representative temporally varying seasonal 1-hour SO 2 background values to determine the overall impact. The resulting concentrations were compared

The COS Vent Oxidizer was the worst-case modeling scenario and was limited to 415 lbs of

SO 2/hr which equates to a 269.21 lbs of SO 2/hr over a 24-hour averaging period. The COS Vent Oxidizer represented 93 percent of SABIC’s total SO2 modeled contributions. The other 7 percent of the modeled contributions were from SABIC’s ancillary units, which also have SO 2 limits, as well as impacts from all other modeled inventory sources. Table 11.5 shows the

modeled results used to establish SABIC’s SO 2 emission limits. The overall maximum concentration was 191.9 µg/m 3; occurring at UTM coordinates: 418467.1 East, 4195409.8 North.

Page 20 of 69

Table 11.5 – SABIC Modeling Results Maximum Modeled 1-Hour SO Facility Concentration 2 NAAQS Models Emission Scenarios Including Seasonal Hourly (µg/m 3) Attainment Background (µg/m 3) SABIC COS Flare 135.4 196.2 Yes SABIC COS Vent Oxidizer 191.9 196.2 Yes

th The concentration isopleths showing the maximum predicted 99 percentile daily 1-hour SO 2 concentration gradients can be found in Figure 11.5. The modeling demonstrated attainment of the 1-hour SO 2 standard with the emission limits listed in SABIC’s Commissioner’s Order.

Figure 11.5 – SABIC Modeling Results

Page 21 of 69

12.0 Lake County: Source IDs ArcelorMittal – USA (18-089- 00316)/Cokenergy (18-089-00383)/U.S. Steel (18-089-00121)

12.1 Source Description ArcelorMittal - USA is an integrated steel mill consisting of two blast furnaces, one sinter plant, one basic oxygen furnace (BOF) complex, one hot metal Reladle/Desulf complex, an 84 inch hot strip mill with three reheat furnaces, mill finishing and sheet finishing operations, plate mill furnaces, two coke batteries, and five power station boilers. Some processes such as the BOF steel making processes have roof monitor emissions in addition to stack emissions. The blast furnaces also have non-point slag pit loadout fugitive emissions which are modeled as volume sources.

Cokenergy is an integrated steel mill consisting of one lime spray dryer Flue Gas Desulfurization unit and baghouse for the heat recovery coal carbonization facility (HRCC) waste gas stream operated by Indiana Harbor Coke Company (IHCC).

U.S. Steel is an integrated steel mill consisting of three coke batteries, a coke plant by-product recovery plant, one coke oven gas desulfurization facility, a coke plant boiler house, a sinter plant, four blast furnaces, two Basic Oxygen Process (BOP) shops with hot metal transfer and desulfurization stations, an 84 inch hot strip mill, a boiler house, and a TurboBlower boiler house. Some processes such as the BOF steel making processes have roof monitor emissions in addition to stack emissions. The blast furnaces also have non-point slag pit fugitive emissions which are modeled as volume sources.

The modeling option was chosen to address the DRR for each of the three DRR sources in Lake County.

12.2 Characterization of Modeled Area ArcelorMittal - USA is located at 3001 Dickey Road, East Chicago, in North Township, Lake County, Indiana. The northern end of the ArcelorMittal plant borders the southern shoreline of Lake Michigan.

Cokenergy is located at 3210 Watling Street, East Chicago, in North Township, Lake County, Indiana. CokeEnergy is located on the same property as ArcelorMittal – USA.

U.S. Steel is located at 1 North Broadway, Gary, in Calumet Township, Lake County, Indiana. The northern end of the U.S. Steel plant borders the southern shoreline of Lake Michigan.

Page 22 of 69

A map of the area surrounding the three DRR facilities in Lake County and the townships in which wach DRR facility is located is shown in Figure 12.1

Figure 12.1 - Lake County DRR Sources and Surrounding Area

12.3 Background Concentrations st The nearest 1-hour SO 2 monitored concentrations were taken from the Hammond 141 Street (AQS #18-089-2008) and Gary-IITRI (AQS #18-089-0022) monitors. The Hammond monitor was used for the western half of the receptor grid and Gary-ITTRI for the eastern half. The 99 th percentile values from 2013 through 2015 and the 3-year design value are listed below in Table 12.1

Page 23 of 69

th Table 12.1 – Lake County 99 Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) Monitoring Site 2013 2014 2015 2013-2015 Hammond 141 st St 23.7 20.2 26.0 a 23 Gary-IITRI 43.2 53.1 35.0 44 a Incomplete data.

12.4 Modeling Methodology The Lake County DRR modeling methodology resembles modeling used to evaluate New Source Review (NSR) and Prevention of Significant Deterioration (PSD) sources. However, Indiana has relied on U.S. EPA guidance “SO 2 NAAQS Designations Modeling Technical Assistance Document” in order to conduct an appropriate air dispersion modeling analysis for Lake County

to support 1-hour SO 2 designation recommendations.

12.4.1 Model Selection In accordance with Appendix A of Appendix W to 40 Code of Federal Regulations (CFR) Part 51, Indiana used the American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD) version 15181. BPIPPRIME was used to account for any building downwash concerns.

12.4.2 Model Options ArcelorMittal - USA/Cokenergy/U.S. Steel used the adjustment to the surface friction velocity, (ADJ_U*), AERMET option in their modeling analysis. This option was recently accepted as a regulatory option in the final rule “Revisions to the Guidelines on Air Quality Models, Enhancements to the AERMOD Dispersion Modeling System and Incorporation of Approaches to Address Ozone and Fine Particulate Matter”, signed on December 20, 2016. The ADJ_U* regulatory option provides for better model performance.

Non-regulatory options within AERMOD were used to determine the air quality characteristics for Lake County. This is due to the use of site-specific meteorology. The area is considered primarily urban, based on population density. The population value used was equal to the sum of population of cities where sources exist and any adjacent cities which meet the population density criteria. Technically, Gary, Indiana did not meet the strict definition of population density for urban classification. However, at least one-quarter of the area of Gary consists of U.S. Steel. By definition an integrated steel mill is considered urban with light-moderate to heavy industrial use. The entire population lives in the remainder of Gary. After factoring out 25% of the Gary’s land area, Gary meets the 750 people/sq km population density threshold for using an urban dispersion coefficient. Therefore, an urban classification with an area population Page 24 of 69

of 243,149 was used in the model input, as provided for in the Guideline on Air Quality Models, Section 7.2.3 (EPA, 2005b). Table 12.2 details the surrounding sizes and population densities of towns in the area to determine the overall population density for the appropriate urban land use characterization. All other regulatory default options were selected to perform the air quality analysis for the three Lake County DRR facilities.

Table 12.2 - Lake County Urban Population Population Population Adjusted Area City Population Density Density Density sq mi per sq mi per sq km per sq km Gary 80,294 49.87 1,610 613 818 Hammond 80,830 22.78 3,548 1,344 N/A East Chicago 29,698 14.09 2,108 950 N/A Whiting 4,997 1.8 2,776 1,081 N/A Munster 23,603 7.57 3,118 1,198 N/A Highland, IN 23,727 6.94 3,419 1,318 N/A

Total 243,149

12.4.3 AERMAP The AERMOD terrain preprocessor mapping program, AERMAP, was used to determine all the terrain elevation heights for each receptor, building, and source locations using the Universal Transverse Mercator (UTM) coordinate system. The most recent AERMAP version 11103 assigned the elevations from the National Elevation Dataset (NED) using the North American Datum (NAD) 1983 as recommended in the, “40 CFR Part 51, Revision to the Guideline on Air Quality Models” Appendix W and later revised in the “AERMOD Implementation Guide.”

12.5 Meteorological Data 12.5.1 AERMET The Gary-IITRI surface meteorological data and the Lincoln, Illinois upper air meteorological data taken from 2013 through 2015 were used to determine the meteorological conditions surrounding the three Lake County DRR sources. The Gary-IITRI surface meteorological data was used to more accurately include the influence of Lake Michigan on the meteorological conditions in the area immediately surrounding the three Lake County DRR facilities. The Gary- IITRI surface data was processed without turbulence parameters in order to use the ADJ_U* option. This was processed with the latest version of the AERMOD meteorological data Page 25 of 69

preprocessor program AERMET (version 15181). Table 12.3 below lists the surface and upper air meteorological stations used to conduct modeling.

Table 12.3 – Lake County NWS Stations/Onsite Meteorological Stations Facility Surface Meteorology Upper Air Meteorology ArcelorMittal-USA/U.S. Steel/ Gary-IITRI Monitor/ Lincoln, IL NWS Cokenergy South Bend NWS

12.5.2 Wind Rose The Gary-IITRI surface meteorological data and the Lincoln, Illinois upper air meteorological data taken from 2013 through 2015 were used to determine the meteorological conditions for the Lake County area. The Gary-IITRI wind rose for the 3-year modeled period 2013-2015 is shown as Figure 12.2 below. The Gary wind rose depicts the predominant wind direction as from the southwest for the 3-year modeled period.

Figure 12.2 – Gary-IITRI 3-year Cumulative Wind Rose (2013 – 2015)

12.5.3 AERMINUTE/AERSURFACE The 1-minute wind speeds and wind directions, taken from the Automated Surface Observing System (ASOS) NWS stations and onsite meteorological stations, were processed with the U.S. EPA 1-minute data processor program AERMINUTE version 15272. All regulatory default options were selected with the exception of the use of the adjustment to the surface friction Page 26 of 69

velocity, (ADJ_U*) option. The ADJ_U* option has been demonstrated to provide better model performance for determining 1-hour SO 2 concentrations. The ADJ_U* option has been accepted by U.S. EPA in a final rulemaking signed on December 20, 2016.

The U.S. EPA program AERSURFACE version 13016 was used to determine the surface characteristics; albedo, Bowen ratio, and surface roughness for the Gary-IITRI, Indiana meteorological tower location. Surface characteristics were determined at the NWS location for each of 12 wind direction sectors with a recommended default radius of one kilometer.

12.6 Receptor Grid and Modeling Domain The receptor grid and modeling domain was based on guidance provided in the memorandum “Updated Guidance for Area Designations for the 2010 Primary Sulfur Dioxide National

Ambient Air Quality Standards”, dated March 20, 2015 and the SO 2 NAAQS Designations Modeling TAD. Indiana used a multi-nested rectangular receptor grid with appropriate spacing of receptors based on the distance from the modeled emission points to detect significant concentration gradients. The modeling domain extended out to include all sources and the

appropriate distances to model maximum 1-hour SO 2 impacts to determine attainment designations for the area. Indiana used the following multi-nested rectangular receptor grids listed below and depicted in Figure 12.3. Focus was emphasized on receptor placement near

each of the Lake County DRR sources; expected 1-hour SO 2 impacts would be anticipated to be very near each source.

• Receptor spacing at the fence line for each facility was placed every 50 meters. • Receptor spacing at 100 meters was placed out to a distance of 5,000 meters (5 kilometers) beyond each facility. • Receptor spacing at 500 meters was placed out to a distance of 10,000 meters (10 kilometers) beyond each facility and east to the Porter County line. • Receptor spacing at 1000 meters was placed beyond 10,000 meters (10 kilometers) from each facility to the south to cover the southern extent of St. John, Ross and North townships. Page 27 of 69

Figure 12.3 – Lake County Receptor Grid

ArcelorMittal - USA, Cokenergy and U.S. Steel have fenced areas, natural boundaries and gated areas with regular security patrols to keep unauthorized people off the property. Since this is the case, receptors were placed along the property lines as appropriate.

12.7 Stack Heights The use of actual stack heights rather than relying on Good Engineering Practice (GEP) stack heights when modeling actual emissions was utilized in the analysis per the SO 2 NAAQS Designations Modeling TAD.

12.8 Temporally Varying Seasonal 1-Hour SO 2 Background

Page 28 of 69

area. The latest three years of SO 2 air quality monitoring data (2013-2015) was used from both the Hammond and Gary sites.

Temporally varying seasonal 1-hour SO 2 background concentrations were taken from the Hammond (west) and Gary (east) monitors for 2013 - 2015. Two sets of 1-hour SO 2 background were used to best represent the Lake County DRR sources, ArcelorMittal – USA and Cokenergy are located in the western portion of the county and U.S. Steel is located in the eastern portion of

the county. Hammond monitor will also measure the SO 2 impacts from Illinois. The hourly seasonal SO 2 values used for representative background concentrations for the Lake County DRR sources are listed below in Table 12.4 for the Hammond monitor and in Table 12.5 for the Gary-IITRI monitor.

Table 12.4 – Lake County Hammond Monitor 99 th Percentile Temporally Varying

Seasonal SO 2 Background Values (ppb) Hr 1 Hr 2 Hr 3 Hr 4 Hr 5 Hr 6 Hr 7 Hr 8

Winter 5.4 5.7 5.94 6.08 6.12 6.18 5.8 6.14 Spring 5.74 5.53 5.44 5.34 5.6 6.07 6.4 7.03 Summer 4.87 4.63 4.6 4.8 5.57 5.28 6.01 6.57 Fall 5.03 4.13 5.34 3.84 4.61 6.35 6.1 6.28

Hr 9 Hr 10 Hr 11 Hr 12 Hr 13 Hr 14 Hr 15 Hr 16

Winter 6.73 7.03 8.76 7.72 7.89 7.18 8.78 7.84 Spring 8.27 8.43 9.19 7.68 8.2 8.09 8.14 8.86 Summer 8.97 7.54 8.77 8.31 9 7.96 8.95 6.51 Fall 8.1 8.04 8.11 6.84 8.08 7.52 8.16 7.74

Hr 17 Hr 18 Hr 19 Hr 20 Hr 21 Hr 22 Hr 23 Hr 24 Winter 6.9 6.18 6.44 5.74 5.58 5.74 5.68 5.58 Spring 8.85 9.4 9.24 7.76 7.9 6.84 7 7.84 Summer 7.76 7.87 7.97 6.31 6.04 8.07 5.69 5.14 Fall 8.91 6.81 7.12 7.31 6.75 5.37 4.9 3.8

Page 29 of 69

Table 12.5 – Lake County Gary - IITRI 99 th Percentiles Temporally Varying

Seasonal SO 2 Background Values (ppb) Hr 1 Hr 2 Hr 3 Hr 4 Hr 5 Hr 6 Hr 7 Hr 8

Winter 9.69 7.35 7.1 6.74 6.87 7.03 6.32 7.42 Spring 7.31 4.59 7.82 4.88 6.88 7.84 8.58 6.96 Summer 1.37 1 1 1 1 1 1 1 Fall 6.98 5.64 5.44 5.56 7.57 4.64 5.24 8.02

Hr 9 Hr 10 Hr 11 Hr 12 Hr 13 Hr 14 Hr 15 Hr 16

Winter 8.35 9.35 9.52 9.35 8.66 8.5 12.29 10.44 Spring 8.22 8.17 10.34 15.5 9.62 9.02 9.54 9.05 Summer 5.83 9.03 7.29 7.47 5.47 4.47 3.93 3.77 Fall 6.9 6.81 8.5 8.82 8.84 8.96 7 6.45

Hr 17 Hr 18 Hr 19 Hr 20 Hr 21 Hr 22 Hr 23 Hr 24 Winter 9.33 6.84 7.22 8.35 6.4 6.81 8.64 9.04 Spring 8.24 7.84 7.38 6.34 7.32 6.44 8.73 7.58 Summer 3.72 3.97 2.53 2.41 2.4 1 2.24 2.83 Fall 6.46 4.62 4.71 7.14 4.64 4.94 7.01 7.19

12.9 SO 2 Emissions Included in the Modeling Analysis 12.9.1 DRR Source Emissions ArcelorMittal - USA and U.S. Steel were modeled using different emission methodologies. Continuous emission monitoring data (CEM) data was available for several emission units while others had seasonal or weekly varying emission rates that were modeled. Cokenergy has emission data collected by a continuous emission monitor; therefore, CEM data was modeled. ArcelorMittal – USA and U.S. Steel have processes with varying hourly emissions rates that were based on a daily maximum emission rate. Emissions were allocated for each hour of the day. Emission units without CEM data or daily emission records were averaged across the three modeled years (2013-2015). Enclosure 2 contains a listing of all of the AERMOD inputs of all the DRR and inventory sources for Lake County.

12.9.2 Carmeuse Lime’s Commissioner’s Order – SO 2 Emission Limits Carmeuse Lime, Inc. (Carmeuse) is a stationary lime manufacturing plant (Source I.D. 089- 00112) located at 1 North Carmeuse Drive in Gary in Lake County. Carmeuse is not a DRR

source but was identified as potentially impacting SO 2 air quality near the Lake County DRR sources. SO 2 sources from the surrounding area in Lake County were evaluated to determine if their emissions would impact the air quality surrounding the DRR sources, beyond what is Page 30 of 69

captured through background SO 2 ambient air monitoring data. Initial modeling, using actual emissions data from Carmeuse showed potential 1-hour SO 2 concentrations higher than the 1- hour SO 2 NAAQS. Therefore, Carmeuse submitted a request on November 15, 2016 for a Commissioner’s Order to establish SO 2 emission limits that would be federally enforceable and permanent which demonstrate attainment of the 1-hour SO 2 standard. The Commissioner’s Order #2016-04 was signed on November 16, 2016 and is included in Enclosure 3.

Carmeuse’s SO 2 emissions are distributed amongst their five kilns. In order to establish hourly emissions limits for Carmeuse through the Commissioner’s Order, modeling was conducted to determine limits that demonstrated compliance with the 1-hour SO 2 standard. Each kiln has six stacks so modeling determined each kiln would be limited to 12.0 pounds of SO 2/hour or 2.0 pounds of SO 2/hour for each stack of each kiln. The three DRR sources, surrounding SO 2 source inventories, and temporally varying seasonal SO 2 background concentrations were included in the modeling to establish Carmeuse’s emission limits through a Commissioner’s Order.

The 720 operating hour rolling average emission limit listed in the Commissioner’s Order was based on the 12.0 pound/hour limit modeled for each kiln. U.S. EPA recommended using a flat averaging ratio for emission units with no emission controls, as referenced in Table 1 of U.S.

EPA’s “Guidance for 1-hour SO 2 Nonattainment Area SIP Submissions”. Based on the average th th ratio of 99 percentile 30-day average SO 2 emission values to the 99 percentile of hourly SO 2 emission values of 0.79, the corresponding 720 operating hour average for each kiln was calculated to be 9.48 lb/hr.

12.9.3 Inventoried SO 2 Sources Included in the Modeling Inclusion of sources in the DRR modeling was based upon their actual emissions from 2013-

2015. The only exception was BP Products (BP), which modeled 2015 SO 2 emissions. BP completed its Whiting Refinery Modernization Project (WRMP) on May 10, 2014. This project was permitted with a significant source modification (Permit #089-25484-00453 issued May 1, 2008) and significant permit modification (Permit #089-25488-00453 issued June 16, 2008), authorizing the construction of new emission units, modifications to existing emission units and operational changes as necessary. A Consent Decree (Civil No. 2:12-cv-00207) was issued to address revisions to BP’s WRMP. SO 2 emissions as a result of the WRMP were modeled for the Lake County DRR analysis.

All facilities greater than one-half of the PSD significance threshold of 40 tpy were included.

The sources which were explicitly modeled had overall SO 2 emissions of 16,233 tpy. This accounts for 99.8% of the Lake County SO 2 inventory. Continuous emissions monitoring data, seasonal or daily varying emissions or an average of 3-year annual SO 2 emissions were modeled for all sources.

Page 31 of 69

The modeled inventory included two Porter County SO2 sources (ArcelorMittal – Burns Harbor and the NIPSCO – Bailly Generating Station). Koppers Inc. in Chicago, Illinois, was also included in the inventory. Two coal-fired power plants in Cook County, Illinois shut down in 2012 and as a result were not included in the modeling analysis. The following facilities were

included in the air quality modeling analysis to determine the overall SO 2 air quality impact in the area and are listed in Table 12.6.

Table 12.6 - Lake County Modeling Inventory 2013-2015 Average Source Source ID Location SO 2 Emissions (tpy) BP Products, North America Inc. 18-089-00003 Lake County, IN 400.2 a Carmeuse Lime, Inc 18-089-00112 Lake County, IN Emission Limits b Eco Services Corp 18-089-00242 Lake County, IN 255.6 Safety-Kleen Systems Inc. 18-089-00301 Lake County, IN 62.6 ArcelorMittal - USA 18-089-00318 Lake County, IN 1,430.8 Indiana Harbor Coke Company 18-089-00382 Lake County, IN 2,441.1 Ironside Energy LLC 18-089-00448 Lake County, IN 204.5 ISPAT Inland LaFarge NA 18-089-00458 Lake County, IN 122.9 ArcelorMittal – Burns Harbor 18-127-00001 Porter County, IN 12,189 NIPSCO Bailly Generating 2013-2015 18-127-00002 Porter County, IN Station CEMS Data Koppers Inc. 170000035076 Cook County, IL 1,785.7 a IDEM utilized BP Products’ 2015 SO 2 emissions due to the Whiting Refinery Modernization Project, completed on May 10, 2014 b Carmeuse Lime, Inc. established SO 2 emission limits in Commissioner’s Order #2016-04

12.10 Modeling Results The 99 th percentile of the 1-hour daily maximum modeled concentrations represents the fourth

to the 1-hour SO 2 standard to indicate whether a modeled violation of the SO 2 NAAQS occurred. All concentrations fell below the 1-hour SO 2 NAAQS and were determined to attain the standard and the area surrounding the DRR sources is recommended as attainment. Table 12.6 shows the modeled localized peaks for all DRR sources in Lake County and including the Carmeuse’s SO 2 emission limits established through the Commissioner’s Order. The overall maximum concentration was 192.2 µg/m 3, occurring at UTM coordinates 466100.0 East, 4609900.0 North, Page 32 of 69

associated with Carmeuse’s maximum impacts. 1-hour SO 2 impacts east of Lake County are being addressed through the air quality characterization of Porter County using the monitoring option for ArcelorMittal – Burns Harbor facility, a DRR source.

Table 12.7 – Lake County Modeling Results Maximum Modeled Concentration 1-Hour SO 2 Models Source Including Seasonal Hourly NAAQS Attainment Background (µg/m 3) (µg/m 3) Carmeuse Lime 192.2 196.2 Yes U.S. Steel 128.1 196.2 Yes Cokenergy 182.8 196.2 Yes ArcelorMittal USA 182.8 196.2 Yes Porter County Line 168.7 196.2 Yes

th The concentration isopleths showing the maximum predicted 99 percentile daily 1-hour SO 2 concentration gradients can be found in Figure 12.5.

Figure 12.4 – Lake County Modeling Results

Page 33 of 69

13.0 Duke-Gallagher (Source ID 153-00005)

13.1 Source Description Duke - Gallagher Generating Station (Duke - Gallagher) is a 280 MW coal-fired power plant in Floyd County located in southeast Indiana. Duke - Gallagher has two coal-fired boilers rated at 1,390 MMBtu/hr each. The plant is operated by Duke Energy Indiana, LLC. Duke - Gallagher was identified as a Data Requirements Rule (DRR) source based on their actual 2014 SO 2 emissions of 3,524 tons exceeding the DRR threshold of 2,000 tons of SO 2.

13.2 Characterization of Modeled Area Duke - Gallagher is located at 30 Jackson St, New Albany, Indiana, on the banks of the Ohio River in New Albany Township, Floyd County, Indiana. A map of the area surrounding Duke - Gallagher is shown below in Figure 13.1.

Figure 13.1 – Duke - Gallagher and Surrounding Area

Page 34 of 69

13.3 Background Concentrations

The nearest 1-hour SO 2 monitored concentrations were taken from the Green Valley monitor (AQS #18-043-1004) located in Floyd County. The 99th percentile values from 2013 through 2015 and the 3-year design value are listed below in Table 13.1.

th Table 13.1 – Duke – Gallagher 99 Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) Monitoring Site 2013 2014 2015 2013-2015 Floyd Co – Green Valley 30.0 65.0 28.0 41

13.4 Modeling Methodology The Duke - Gallagher DRR modeling methodology resembles modeling used to evaluate New Source Review (NSR) and Prevention of Significant Deterioration (PSD) sources. However,

Indiana has relied on U.S. EPA guidance “SO 2 NAAQS Designations Modeling Technical Assistance Document” in order to conduct an appropriate air dispersion modeling analysis for

Duke - Gallagher to support 1-hour SO 2 designation recommendations.

13.4.1 Model Selection In accordance with Appendix A of Appendix W to 40 Code of Federal Regulations (CFR) Part 51, Indiana used the American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD) version 15181. BPIPPRIME was used to account for any building downwash concerns.

13.4.2 Model Options All regulatory default options within AERMOD were used to determine the air quality characteristics surrounding Duke - Gallagher. The area is considered primarily rural, based on the Auer’s Classification Land Use methodology with a vast majority of the land use types classified as undeveloped rural (A4), water surfaces (A5) and estate residential (R4). Therefore, a rural classification was used, as provided for in the Guideline on Air Quality Models, Section 7.2.3 (EPA, 2005b). No variation of the population selection was necessary. Figure 13.2 shows the 3-kilometer radius area surrounding Duke - Gallagher that was analyzed to determine the land use classification.

Page 35 of 69

Figure 13.2 – Duke – Gallagher 3-kilometer Radius to Determine Auer Land Use

13.4.3 AERMAP The AERMOD terrain preprocessor mapping program, AERMAP, was used to determine all the terrain elevation heights for each receptor, building, and source locations using the Universal Transverse Mercator (UTM) coordinate system. The most recent AERMAP version 11103 assigned the elevations from the National Elevation Dataset (NED) using the North American Datum (NAD) 1983 as recommended in the, “40 CFR Part 51, Revision to the Guideline on Air Quality Models” Appendix W and later revised in the “AERMOD Implementation Guide.”

13.5 Meteorological Data 13.5.1 AERMET

Page 36 of 69

preprocessor program AERMET (version 15181). Table 13.2 below lists surface and upper air meteorological stations used to conduct modeling.

Table 13.2 – Duke - Gallagher NWS Stations/Onsite Meteorological Stations Facility Surface Meteorology Upper Air Meteorology Duke – Gallagher Louisville, KY NWS Wilmington, OH NWS

13.5.2 Wind Rose The Louisville, Kentucky National Weather Service (NWS) surface meteorological data and Wilmington, Ohio upper air meteorological data taken from 2013 through 2015 were used to determine the meteorological conditions for the area surrounding Duke - Gallagher in AERMOD. The Louisville NWS wind rose for the 3-year modeled period 2013-2015 is shown as Figure 13.3 below. The Louisville NWS wind rose depicts the predominant wind direction as from the southwest for the 3-year modeled period 2013-2015.

Figure 13.3 - Louisville 3-year Cumulative Wind Rose (2013 – 2015)

13.5.3 AERMINUTE/AERSURFACE The 1-minute wind speeds and wind directions, taken from the Automated Surface Observing System (ASOS) NWS stations and onsite meteorological stations, were processed with the U.S. EPA 1-minute data processor program AERMINUTE version 15272. Page 37 of 69

The U.S. EPA program AERSURFACE version 13016 was used to determine the surface characteristics; albedo, Bowen ratio, and surface roughness for the Louisville, Indiana NWS meteorological tower location. Surface characteristics were determined at the NWS location for each of 12 wind direction sectors with a recommended default radius of one kilometer.

13.6 Receptor Grid and Modeling Domain The receptor grid and modeling domain was based on guidance provided in the memorandum “Updated Guidance for Area Designations for the 2010 Primary Sulfur Dioxide National

Ambient Air Quality Standards”, dated March 20, 2015 and the SO 2 NAAQS Designations Modeling TAD. Indiana used a multi-nested rectangular receptor grid with appropriate spacing of receptors based on the distance from the modeled emission points to detect significant concentration gradients. The modeling domain extended out to include all sources and the

• Receptor spacing at the fence line for each facility was placed every 50 meters. • Receptor spacing at 100 meters was placed out to a distance of 3,000 meters (3 kilometers) beyond the DRR facility. • Receptor spacing at 250 meters was placed out to a distance of 5,000 meters (5 kilometers) beyond the DRR facility. • Receptor spacing at 500 meters was placed out to a distance of 10,000 meters (10 kilometers) beyond the DRR facility.

Page 38 of 69

Figure 13.4 – Duke – Gallagher Receptor Grid

Duke - Gallagher has a fenceline, natural features, and security patrols that restrict public access to its property. Receptors were therefore placed along the property boundary where public access is not restricted.

13.7 Stack Heights The use of actual stack heights rather than relying on Good Engineering Practice (GEP) stack heights when modeling actual emissions was utilized in the analysis per the SO 2 NAAQS Designations Modeling TAD.

13.8 Temporally Varying Seasonal 1-Hour SO 2 Background

Temporally varying seasonal 1-hour SO 2 background concentrations were taken from the Green Valley monitor (AQS #18-043-1004) located in Floyd County for 2013 - 2015. The hourly

seasonal SO 2 values used for representative background concentrations for the area surrounding Duke - Gallagher are listed below in Table 13.3.

Table 13.3 – Duke – Gallagher 99 th Percentile Temporally Varying

Seasonal SO 2 Background Values (ppb) Hr 1 Hr 2 Hr 3 Hr 4 Hr 5 Hr 6 Hr 7 Hr 8

Winter 7.27 6.90 6.40 5.80 5.82 6.69 4.36 7.85 Spring 8.01 7.38 4.23 7.32 4.86 3.90 4.28 6.25 Summer 5.60 3.46 4.10 3.47 2.57 1.89 2.30 3.70 Fall 3.70 3.76 4.23 4.06 3.13 3.30 6.33 7.51

Hr 9 Hr 10 Hr 11 Hr 12 Hr 13 Hr 14 Hr 15 Hr 16

Winter 7.24 9.10 8.98 10.66 9.42 6.60 9.96 9.70 Spring 8.39 8.87 9.50 16.88 13.04 15.89 9.10 14.09 Summer 7.70 8.10 13.52 13.08 13.15 8.94 8.57 7.78 Fall 6.96 9.52 9.46 8.82 8.87 9.06 13.28 8.62

Hr 17 Hr 18 Hr 19 Hr 20 Hr 21 Hr 22 Hr 23 Hr 24 Winter 10.21 9.54 8.78 8.45 7.77 8.32 7.92 6.43 Spring 15.33 9.21 9.63 9.94 8.06 7.24 7.70 8.15 Summer 6.22 8.08 6.56 4.87 3.73 3.47 4.16 3.46 Fall 11.71 6.29 6.93 6.42 5.47 3.60 3.53 5.31

13.9 SO 2 Emissions Included in the Modeling Analysis

13.9.1 DRR Source: Duke - Gallagher Emissions Duke - Gallagher has two coal-fired units, Units 2 and 4 that have continuous emission monitoring (CEM) data for SO 2. This hourly CEM data from both units was formatted and used in the 1-hour SO 2 AERMOD model run.

Page 40 of 69

13.9.2 Inventoried SO 2 Sources Included in the Modeling

SO 2 sources from the surrounding area were evaluated to determine if their SO 2 emissions had a potential impact on the air quality surrounding the DRR source, beyond what is captured through background monitoring data. The average actual emissions from 2013-2015 were input for ESSROC and Louisville Medical Center Steam Plant. Louisville Gas & Electric facilities at

Cane Run and Mill Creek have reduced their SO 2 emissions with federal regulatory measures including the Mercury and Air Toxics rule, Cross State Air Pollution rule and several other federal rule-makings. SO2 emission reductions will be achieved through conversion of the coal- fired electric generating units to a natural gas combined cycle unit for Cane Run and additional

SO 2 flue-gas desulfurization (FGD) controls and upgrades at the Mill Creek facility. Permitted limits were modeled for each of the Louisville Gas and Electric sources as the emission reductions are federally enforceable and permanent. The following list of sources, found below in Table 13.4, were included in the AERMOD run to determine overall air quality characteristics.

Table 13.4 – Duke – Gallagher Modeling Source Inventory 2013-2015

Source Source ID Location SO 2 Emissions (tpy) ESSROC Cement Corporation 18-019-00008 Clark County, IN 416 LG & E – Cane Run 21-111-00126 Jefferson County, KY 21 LG & E – Mill Creek 21-111-00127 Jefferson County, KY 13,485 Louisville Medical Center 21-111-00148 Jefferson County, KY 415

13.10 Modeling Results The 99 th percentile of the 1-hour daily maximum modeled concentrations represents the fourth

percentile daily 1-hour SO 2 concentration is shown in Table 13.5. The overall maximum concentration was 99.5 µg/m 3, occurring at UTM coordinates 602300.0 East, 4238000.0 North.

Page 41 of 69

Table 13.5 - Duke – Gallagher Modeling Results Maximum 1-Hour SO Facility Modeled Concentration 2 Emission Scenario NAAQS Models Including Seasonal Hourly (µg/m 3) Attainment Background (µg/m 3) Gallagher 99.5 196.2 Yes

th The concentration isopleths showing the maximum predicted 99 percentile daily 1-hour SO 2 concentration gradients can be found in Figure 13.5.

Figure 13.5 – Duke - Gallagher Modeling Results

Page 42 of 69

14.0 NIPSCO – R.M. Schahfer Generating Station (Source ID 18-073-00008)

14.1 Source Description The Northern Indiana Public Service Company (NIPSCO) - R.M. Schahfer Generating Station (NIPSCO - Schahfer) is a stationary electric utility generating station consisting of four units that have a capacity to generate 1,943 megawatts (MW) of electricity combined. NIPSCO - Schahfer has four coal-fired boilers; one boiler is rated at 4,650 MMBtu/hr, one boiler is rated at 5,100 MMBtu/hr, and two boilers are rated at 3,967 MMBtu/hr each. The plant is operated by NiSource.

14.2 Characterization of Modeled Area The NIPSCO - Schahfer is located at 2723 East 1500 North, Wheatfield, in Kankakee Township, Jasper County, Indiana; approximately 5 miles west of State Road 421. A map of the area surrounding the NIPSCO - Schahfer facility is shown below in Figure 14.1.

Figure 14.1 - NIPSCO - Schahfer and Surrounding Area

Page 43 of 69

14.3 Background Concentrations

The nearest 1-hour SO 2 monitored concentrations were taken from the Wheatfield – Jasper County monitor (AQS #18-073-0002). The 99 th percentile values from 2012 through 2014 and the 3-year design value are listed below in Table 14.1.

th Table 14.1 – NIPSCO – Schahfer 99 Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) Monitoring Site 2012 2013 2014 2012-2014 Wheatfield – Jasper County 33 40 18 30

14.4 Modeling Methodology The NIPSCO - Schahfer DRR modeling methodology resembles modeling used to evaluate New Source Review (NSR) and Prevention of Significant Deterioration (PSD) sources. However,

Indiana has relied on U.S. EPA guidance “SO 2 NAAQS Designations Modeling Technical Assistance Document” in order to conduct an appropriate air dispersion modeling analysis for

NIPSCO - Schahfer to support 1-hour SO 2 designation recommendations.

14.4.1 Model Selection In accordance with Appendix A of Appendix W to 40 Code of Federal Regulations (CFR) Part 51, Indiana used the American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD) version 15181. BPIPPRIME was used to account for any building downwash concerns.

14.4.2 Model Options All regulatory default options within AERMOD were used to determine the air quality characteristics surrounding NIPSCO - Schahfer. The area is considered primarily rural, based on the Auer’s Classification Land Use methodology with a vast majority of the land use types classified as agricultural rural (A2), undeveloped rural (A4), water surfaces (A5) and estate residential (R4). Therefore, a rural classification was used, as provided for in the Guideline on Air Quality Models, Section 7.2.3 (EPA, 2005b). No variation of the population selection was necessary. Figure 14.2 shows the 3-kilometer radius area surrounding NIPSCO - Schahfer that was analyzed to determine the land use classification.

Page 44 of 69

Figure 14.2 – NIPSCO – Schahfer 3-km Radius to Determine Auer Land Use

14.4.3 AERMAP The AERMOD terrain preprocessor mapping program, AERMAP, was used to determine all the terrain elevation heights for each receptor, building, and source locations using the Universal Transverse Mercator (UTM) coordinate system. The most recent AERMAP version 11103 assigned the elevations from the National Elevation Dataset (NED) using the North American Datum (NAD) 1983 as recommended in the, “40 CFR Part 51, Revision to the Guideline on Air Quality Models” Appendix W and later revised in the “AERMOD Implementation Guide.”

14.5 Meteorological Data 14.5.1 AERMET

Page 45 of 69

preprocessor program AERMET (version 15181). Table 14.2 below lists surface and upper air meteorological stations used to conduct modeling.

Table 14.2 – NIPSCO – Schahfer NWS Stations/Onsite Meteorological Stations Facility Surface Meteorology Upper Air Meteorology NIPSCO - Schahfer South Bend, IN NWS Lincoln, IL NWS

14.5.2 Wind Rose The South Bend National Weather Service (NWS) surface meteorological data and the Lincoln, Illinois upper air meteorological data taken from 2012 through 2014 were used to determine the meteorological conditions for the area surrounding NIPSCO - Schahfer in AERMOD. The South Bend NWS wind rose for the 3-year modeled period 2012-2014 is shown as Figure 14.3 below. The South Bend NWS wind rose depicts the predominant wind direction as from the southwest for the 3-year modeled period 2012-2014.

Figure 14.3 - South Bend 3-year Cumulative Wind Rose (2012 – 2014)

14.5.3 AERMINUTE/AERSURFACE The 1-minute wind speeds and wind directions, taken from the Automated Surface Observing System (ASOS) NWS stations and onsite meteorological stations, were processed with the U.S. EPA 1-minute data processor program AERMINUTE version 15272.

Page 46 of 69

The U.S. EPA program AERSURFACE version 13016 was used to determine the surface characteristics; albedo, Bowen ratio, and surface roughness for the South Bend, Indiana NWS meteorological tower location. Surface characteristics were determined at the NWS location for each of 12 wind direction sectors with a recommended default radius of one kilometer.

14.6 Receptor Grid and Modeling Domain The receptor grid and modeling domain was based on guidance provided in the memorandum “Updated Guidance for Area Designations for the 2010 Primary Sulfur Dioxide National

Ambient Air Quality Standards”, dated March 20, 2015 and the SO 2 NAAQS Designations Modeling TAD. Indiana used a multi-nested rectangular receptor grid with appropriate spacing of receptors based on the distance from the modeled emission points to detect significant concentration gradients. The modeling domain extended out to include all sources and the

appropriate distances to model maximum 1-hour SO 2 impacts to determine attainment designations for the area. Indiana used the following multi-nested rectangular receptor grid, which are listed below and depicted in Figure 14.4:

Page 47 of 69

Figure 14.4 – NISPCO - Schahfer Receptor Grid

NIPSCO - Schahfer’s property line is very extensive. Their property is nearly two miles long and is approximately 1.6 miles wide. NIPSCO - Schahfer is largely fenced and has regular security patrols to keep unauthorized people off the property. Since this is the case, receptors were placed along the property lines.

14.7 Stack Heights The use of actual stack heights rather than relying on Good Engineering Practice (GEP) stack heights when modeling actual emissions was utilized in the analysis per the SO 2 NAAQS Designations Modeling TAD.

14.8 Temporally Varying Seasonal 1-Hour SO 2 Background

Temporally varying seasonal SO 2 background concentrations were developed in accordance with the recommended U.S. EPA guidance for establishment of such background concentrations in

Page 48 of 69

Section 8.2 of 40 CFR Part 51, Appendix W and considered appropriate and representative of the area. The latest three years of SO 2 air quality monitoring data (2012-2014) was used.

Temporally varying seasonal 1-hour SO 2 background concentrations were taken from the Wheatfield monitor for 2012 - 2014. The hourly seasonal SO 2 values used for representative background concentrations for the area surrounding NIPSCO - Schahfer are listed below in Table 14.3.

Table 14.3 – NIPSCO – Schahfer 99 th Percentile Temporally Varying

Seasonal SO 2 Background Values (ppb) Hr 1 Hr 2 Hr 3 Hr 4 Hr 5 Hr 6 Hr 7 Hr 8

Winter 4.75 5.00 4.71 4.68 4.00 5.00 5.40 4.00 Spring 5.54 4.57 5.60 6.16 4.55 5.00 4.47 7.00 Summer 2.44 3.43 3.00 3.45 3.00 3.00 3.49 6.53 Fall 5.26 4.00 4.00 4.00 9.00 7.41 5.29 5.49

Hr 9 Hr 10 Hr 11 Hr 12 Hr 13 Hr 14 Hr 15 Hr 16

Winter 5.00 7.00 7.00 7.00 7.64 7.00 7.00 7.00 Spring 9.52 8.53 8.06 8.00 7.57 7.00 7.98 6.71 Summer 10.16 8.63 8.00 8.86 9.00 9.28 7.66 7.00 Fall 9.00 7.00 7.69 7.64 5.00 6.00 6.62 5.62

Hr 17 Hr 18 Hr 19 Hr 20 Hr 21 Hr 22 Hr 23 Hr 24 Winter 7.00 7.00 6.32 5.00 5.68 6.66 6.00 6.00 Spring 5.00 4.66 7.18 7.60 6.57 5.00 4.57 4.55 Summer 4.56 4.54 6.00 7.44 5.00 3.00 3.40 2.52 Fall 5.00 6.18 6.02 5.48 4.00 5.00 4.00 7.99

14.9 SO 2 Emissions Included in the Modeling Analysis 14.9.1 DRR Source: NIPSCO - Schahfer Generating Station Emissions NIPSCO - Schahfer has four units, Units BLR4, BLR15, BLR17, and BLR18 that have continuous emission monitoring (CEM) data for SO 2. This hourly CEM data from the four units were formatted and used in the 1-hour SO 2 AERMOD model run. Total annual emissions from NIPSCO – Schahfer from 2015 are approximately one-eighth of the emissions from 2012

Page 49 of 69

through 2014 emissions. Therefore, modeling the 2012-2014 emissions is conservative in nature and will be used for this analysis.

14.9.2 Inventoried SO 2 Sources Included in the Modeling

SO 2 sources from the surrounding area were evaluated to determine if their SO 2 emissions had a potential impact on the air quality surrounding the DRR source, beyond what is captured through background monitoring data. Saint Joseph’s College was found to be within 30 kilometers of NIPSCO - Schahfer. Saint Joseph’s College is no longer a Title V source. The college’s last emission report was in 2012. Those emissions were used in the modeling analysis for NIPSCO - Schahfer as listed in Table 14.4.

Table 14.4 – NIPSCO – Schahfer Modeling Source Inventory 2012 SO Emissions Source Source ID Location 2 (tpy) St. Joseph College 073-00001 Jasper County 120.5

14.10 Modeling Results The 99 th percentile of the 1-hour daily maximum modeled concentrations represents the fourth

th The maximum predicted 99 percentile daily 1-hour SO 2 concentration is shown in Table 14.5. The overall maximum concentration was 162.7 µg/m 3, occurring at UTM coordinates 499354.6 East, 4561322.6 North.

Table 14.5 – NIPSCO – Schahfer Modeling Results Maximum 1-Hour SO Facility Modeled Concentration 2 NAAQS Models Emission Scenarios Including Seasonal Hourly (µg/m 3) Attainment Background (µg/m 3) NIPSCO - Schahfer 162.7 196.2 Yes

Page 50 of 69

th The concentration isopleths showing the maximum predicted 99 percentile daily 1-hour SO 2 concentration gradients can be found in Figure 14.5. The modeling showed attainment of the 1- hour SO 2 standard.

Figure 14.5 – NIPSCO - Schahfer Modeling Results

15.0 Hoosier Energy - Merom (Source ID 153-00005)

15.1 Source Description Hoosier Energy - Merom Generating Station (Hoosier Energy - Merom) is a 1070 MW coal fired power plant located in Sullivan County in Southwest Indiana. Hoosier Energy - Merom operates two coal-fired boilers each rated at 5,088 mmBtu/hr. SO 2 emission controls at the facility include a flue gas desulfurization system. Hoosier Energy - Merom was identified as a Data

Requirements Rule (DRR) source based on their actual 2014 SO 2 emissions of 3,318 tons exceeding the DRR threshold of 2,000 tons of SO 2.

Page 51 of 69

15.2 Characterization of Modeled Area Hoosier Energy - Merom is located at 5500 W Old 54, Sullivan, Indiana, approximately 5 miles east of the Wabash River in Gill Township, Sullivan County, Indiana. A map of the area is shown below in Figure 15.1.

Figure 15.1 – Hoosier Energy - Merom and Surrounding Area

15.3 Background Concentrations

The nearest 1-hour SO 2 monitored concentrations were taken from the Terre Haute – Lafayette Road monitor (AQS #18-167-0018). The 99 th percentile values from 2013 through 2015 and the 3-year design value are listed below in Table 15.1. The area surrounding the Lafayette Road

monitor has been addressed through revisions to the 1-hour SO 2 Nonattainment Area State Implementation Plan.

Page 52 of 69

th Table 15.1 – Hoosier Energy – Merom 99 Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) Monitoring Site 2013 2014 2015 2013-2015 Terre Haute – Lafayette Rd 79.1 85.0 71.0 78

15.4 Modeling Methodology The Hoosier Energy - Merom DRR modeling methodology resembles modeling used to evaluate New Source Review (NSR) and Prevention of Significant Deterioration (PSD) sources.

However, Indiana has relied on U.S. EPA guidance “SO 2 NAAQS Designations Modeling Technical Assistance Document” in order to conduct an appropriate air dispersion modeling analysis for Hoosier Energy -Merom to support 1-hour SO 2 designation recommendations.

15.4.1 Model Selection In accordance with Appendix A of Appendix W to 40 Code of Federal Regulations (CFR) Part 51, Indiana used the American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD) version 15181. BPIPPRIME was used to account for any building downwash concerns.

15.4.2 Model Options All regulatory default options within AERMOD were used to determine the air quality characteristics surrounding the Hoosier Energy - Merom. The area is considered primarily rural, based on the Auer’s Classification Land Use methodology with a vast majority of the land use types classified as agricultural rural (A2) and water surfaces (A5). Therefore, a rural classification was used, as provided for in the Guideline on Air Quality Models, Section 7.2.3 (EPA, 2005b). No variation of the population selection was necessary. Figure 15.2 shows the 3- kilometer radius area surrounding Hoosier Energy - Merom that was analyzed to determine the land use classification.

Page 53 of 69

Figure 15.2 – Hoosier Energy – Merom 3-km Radius to Determine Auer Land Use

15.4.3 AERMAP The AERMOD terrain preprocessor mapping program, AERMAP, was used to determine all the terrain elevation heights for each receptor, building, and source locations using the Universal Transverse Mercator (UTM) coordinate system. The most recent AERMAP version 11103 assigned the elevations from the National Elevation Dataset (NED) using the North American Datum (NAD) 1983 as recommended in the, “40 CFR Part 51, Revision to the Guideline on Air Quality Models” Appendix W and later revised in the “AERMOD Implementation Guide.”

15.5 Meteorological Data 15.5.1 AERMET

Page 54 of 69

preprocessor program AERMET (version 15181). Table 15.2 below lists surface and upper air meteorological stations used to conduct modeling.

Table 15.2 – Hoosier Energy – Merom NWS Stations/Onsite Meteorological Stations Facility Surface Meteorology Upper Air Meteorology Hoosier Energy – Merom Evansville, IN NWS Lincoln, IL NWS

15.5.2 Wind Rose The Evansville, Indiana National Weather Service (NWS) surface meteorological data and the Lincoln, Illinois upper air meteorological data taken from 2013 through 2015 were used to determine the meteorological conditions for the area surrounding Hoosier Energy - Merom in AERMOD. The Evansville NWS wind rose for the 3-year modeled period 2013-2015 is shown as Figure 15.3 below. The Evansville NWS wind rose depicts the predominant wind direction as from the southwest for the 3-year modeled period 2013-2015.

Figure 15.3 - Evansville 3-year Cumulative Wind Rose (2013 – 2015)

15.5.3 AERMINUTE/AERSURFACE The 1-minute wind speeds and wind directions, taken from the Automated Surface Observing System (ASOS) NWS stations and onsite meteorological stations, were processed with the U.S. EPA 1-minute data processor program AERMINUTE version 15272.

Page 55 of 69

15.6 Receptor Grid and Modeling Domain The receptor grid and modeling domain was based on guidance provided in the memorandum “Updated Guidance for Area Designations for the 2010 Primary Sulfur Dioxide National

Ambient Air Quality Standards”, dated March 20, 2015 and the SO 2 NAAQS Designations Modeling TAD. Indiana used a multi-nested rectangular receptor grid with appropriate spacing of receptors based on the distance from the modeled emission points to detect significant concentration gradients. The modeling domain extended out to include all sources and the

Page 56 of 69

Figure 15.4 – Hoosier Energy – Merom Receptor Grid

Hoosier Energy - Merom has a fence surrounding the property with security gates restricting public access to all Merom property. Natural barriers immediately surround the property with a reservoir west of the facility and a landfill to the north. Receptors were therefore placed along the property boundary where public access is not restricted.

15.7 Stack Heights The use of actual stack heights rather than relying on Good Engineering Practice (GEP) stack heights when modeling actual emissions was utilized in the analysis per the SO 2 NAAQS Designations Modeling TAD.

15.8 Temporally Varying Seasonal 1-Hour SO 2 Background

Temporally varying seasonal SO 2 background concentrations were developed in accordance with the recommended U.S. EPA guidance for establishment of such background concentrations in

Page 57 of 69

Section 8.2 of 40 CFR Part 51, Appendix W and considered appropriate and representative of the area. The latest three years of SO 2 air quality monitoring data (2013-2015) was used.

Temporally varying seasonal 1-hour SO 2 background concentrations were taken from the Terre Haute – Lafayette Road monitor for 2013 - 2015. The hourly seasonal SO 2 values used for representative background concentrations for the area surrounding Hoosier Energy - Merom are listed below in Table 15.3.

Table 15.3 – Hoosier Energy – Merom 99 th Percentile Temporally Varying

Seasonal SO 2 Background Values (ppb) Hr 1 Hr 2 Hr 3 Hr 4 Hr 5 Hr 6 Hr 7 Hr 8

Winter 4.99 5.61 5.59 5.17 5.56 5.96 6.30 6.69 Spring 5.25 6.70 7.97 4.37 6.82 4.37 5.46 4.78 Summer 2.78 2.54 2.69 2.17 1.81 2.13 2.71 3.81 Fall 8.21 5.06 5.17 4.07 5.87 3.72 3.81 4.35

Hr 9 Hr 10 Hr 11 Hr 12 Hr 13 Hr 14 Hr 15 Hr 16

Winter 6.22 5.45 9.07 11.45 10.06 9.25 7.76 8.97 Spring 6.86 6.29 24.67 11.51 14.16 10.08 6.30 9.29 Summer 4.44 8.83 8.55 10.09 8.43 24.15 26.75 29.68 Fall 6.35 6.03 34.92 18.80 11.22 14.39 7.32 15.27

Hr 17 Hr 18 Hr 19 Hr 20 Hr 21 Hr 22 Hr 23 Hr 24 Winter 10.45 16.58 8.77 8.84 7.05 6.47 8.66 6.99 Spring 8.60 16.86 5.33 4.59 8.55 4.05 5.73 6.31 Summer 12.49 6.59 5.55 3.94 6.82 4.93 4.07 2.74 Fall 5.14 5.22 5.23 5.65 9.28 7.68 9.08 8.03

15.9 SO 2 Emissions Included in the Modeling Analysis 15.9.1 DRR Source: Hoosier Energy - Merom Emissions Hoosier Energy - Merom operates two coal-fired units each of which are equipped with Continuous Emission Monitoring (CEM) systems. CEM data from 2013 through 2015 was formatted into an AERMOD ready hourly input file and used in the final modeling.

Page 58 of 69

15.9.2 Inventoried SO 2 Sources Included in the Modeling

SO 2 sources from the surrounding area were evaluated to determine if their SO 2 emissions had a potential impact on the air quality surrounding Hoosier Energy - Merom, beyond what is captured through background monitoring data. The latest available actual emissions were used for inventory sources. Two sources were included in the model in addition to the Hoosier Energy - Merom facility: Rain II Carbon in Illinois and the Duke - Wabash facility in Vigo County, Indiana.

Rain CII Carbon is a green petroleum coke calcining facility that produces aluminum and other raw materials. Rain CII Carbon is located in Crawford County, Illinois, 20 km southwest of

Hoosier Energy - Merom and produced 3,132 tpy of SO2 in 2014. Hourly continuous emission monitoring data from 2013 through 2015 were used in AERMOD for the Rain II facility.

Duke Energy - Wabash was an electric generating facility in located 51 km to the north of Hoosier Energy - Merom in Vigo County, Indiana. The facility retired all of its coal-fired electric generating units (Units 2-6). Units 2-5 were retired on April 16, 2016 and Unit 6 was retired on December 7, 2016. Although this source was outside of the 30 km radius Indiana used to determine background sources, Indiana included this source in the modeling of Hoosier Energy - Merom due to high background concentrations over the 2013-2015 time period. Upwind impacts in the background data from the Wabash facility were adjusted to prevent double counting. Average actual emissions from 2013 through 2015 was used in the modeling and listed in Table 15.4. Table 15.4 – Hoosier Energy – Merom Modeling Source Inventory

Source Source ID Location SO 2 Emissions (tpy) Rain CII Carbon 033025AAJ Crawford County, IL 2,750 Duke - Wabash 167-00021 Vigo County 28,154

15.10 Modeling Results The 99 th percentile of the 1-hour daily maximum modeled concentrations represents the fourth

to the 1-hour SO 2 standard to indicate whether a modeled violation of the SO 2 NAAQS occurred. All concentrations fell below the 1-hour SO 2 NAAQS and were determined to attain the standard and the area surrounding Hoosier Energy - Merom is recommended as attainment. The th maximum predicted 99 percentile daily 1-hour SO 2 concentration is shown in Table 15.5. The overall maximum concentration was 63.0 µg/m 3, occurring at UTM coordinates 455600.0 East, 4323300.0 North. Page 59 of 69

Table 15.5 – Hoosier Energy – Merom Modeling Results Maximum 1-Hour SO Facility Modeled Concentration 2 NAAQS Models Emission Scenarios Including Seasonal Hourly (µg/m 3) Attainment Background (µg/m 3) Hoosier Energy – Merom 63.0 196.2 Yes

th The concentration isopleths showing the maximum predicted 99 percentile daily 1-hour SO 2 concentration gradients can be found in Figure 15.5.

Figure 15.5 – Hoosier Energy - Merom Modeling Results

Page 60 of 69

16.0 - Duke - Cayuga Generating Station (Source ID 18-165-00001)

16.1 Source Description Duke - Cayuga Generating Station (Duke - Cayuga) is an electric generating station owned by Duke Energy Indiana, LLC. Duke - Cayuga is a two-unit generating facility built between 1967 and 1968. Units 1 and 2 are equipped with scrubbers to reduce the stations sulfur dioxide emissions by approximately 95 percent. The two coal-fired boilers are rated at 4,802 MMBtu/hour each and have a generating capacity of 1104 megawatts. Duke - Cayuga was identified as a Data Requirements Rule (DRR) source based on their actual 2014 SO 2 emissions of 3448.4 tons exceeding the DRR threshold of 2,000 tons of SO 2.

16.2 Characterization of Modeled Area The Duke - Cayuga is located off of State Road 63, Cayuga, Indiana on the banks of the Wabash River, Eugene Township, Vermillion County, Indiana. A map of the area surrounding Duke - Cayuga used for DRR modeling is shown in Figure 16.1.

Figure 16.1 - Duke - Cayuga and Surrounding Area

Page 61 of 69

16.3 Background Concentrations

The nearest 1-hour SO 2 monitored concentrations were taken from the Fountain County monitor (AQS #18-045-0001). The 99 th percentile values from 2012 through 2014 and the 3-year design value are listed below in Table 16.1.

th Table 16.1 – Duke – Cayuga 99 Percentile 1-hour SO 2 Background Values and 3-year Design Value (ppb) Monitoring Site 2012 2013 2014 2012-2014 Fountain County 30 34 22 29

16.4 Modeling Methodology The Duke - Cayuga DRR modeling methodology resembles modeling used to evaluate New Source Review (NSR) and Prevention of Significant Deterioration (PSD) sources. However,

Indiana has relied on U.S. EPA guidance “SO 2 NAAQS Designations Modeling Technical Assistance Document” in order to conduct an appropriate air dispersion modeling analysis for

Duke - Cayuga to support 1-hour SO 2 designation recommendations.

16.4.1 Model Selection In accordance with Appendix A of Appendix W to 40 Code of Federal Regulations (CFR) Part 51, Indiana used the American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD) version 15181. BPIPPRIME was used to account for any building downwash concerns.

16.4.2 Model Options All regulatory default options within AERMOD were used to determine the air quality characteristics surrounding Duke. The area is considered primarily rural, based on the Auer’s Classification Land Use methodology with a vast majority of the land use types classified as agricultural rural (A2), undeveloped rural (A4) and water surfaces (A5). Therefore, a rural classification was used, as provided for in the Guideline on Air Quality Models, Section 7.2.3 (EPA, 2005b). No variation of the population selection was necessary. Figure 16.2 shows the 3- kilometer radius area surrounding Duke - Cayuga that was analyzed to determine the land use classification.

Page 62 of 69

Figure 16.2 – Duke – Cayuga 3-km Radius to Determine Auer Land Use

16.4.3 AERMAP The AERMOD terrain preprocessor mapping program, AERMAP, was used to determine all the terrain elevation heights for each receptor, building, and source locations using the Universal Transverse Mercator (UTM) coordinate system. The most recent AERMAP version 11103 assigned the elevations from the National Elevation Dataset (NED) using the North American Datum (NAD) 1983 as recommended in the, “40 CFR Part 51, Revision to the Guideline on Air Quality Models” Appendix W and later revised in the “AERMOD Implementation Guide.”

16.5 Meteorological Data 16.5.1 AERMET

Page 63 of 69

preprocessor program AERMET (version 15181). Table 16.2 below lists surface and upper air meteorological stations used to conduct modeling

Table 16.2 – Duke – Cayuga NWS Stations/Onsite Meteorological Stations Facility Surface Meteorology Upper Air Meteorology Duke - Cayuga Indianapolis, IN NWS Lincoln, IL NWS

16.5.2 Wind Rose The Indianapolis, Indiana National Weather Service (NWS) surface meteorological data and the Lincoln, Illinois upper air meteorological data taken from 2012 through 2014 was used to determine the meteorological conditions for the area surrounding Duke - Cayuga in AERMOD. The Indianapolis NWS wind rose for the 3-year modeled period 2012-2014 is shown as Figure 16.3 below. The Indianapolis NWS wind rose depicts the predominant wind direction as from the southwest for the 3-year modeled period 2012-2014.

Figure 16.3 – Indianapolis 3-year Cumulative Wind Rose (2012 – 2014)

16.5.3 AERMINUTE/AERSURFACE The 1-minute wind speeds and wind directions, taken from the Automated Surface Observing System (ASOS) NWS stations and onsite meteorological stations, were processed with the U.S. EPA 1-minute data processor program AERMINUTE version 15272.

Page 64 of 69

The U.S. EPA program AERSURFACE version 13016 was used to determine the surface characteristics; albedo, Bowen ratio, and surface roughness for the Indianapolis, Indiana NWS meteorological tower location. Surface characteristics were determined at the NWS location for each of 12 wind direction sectors with a recommended default radius of one kilometer.

16.6 Receptor Grid and Modeling Domain The receptor grid and modeling domain was based on guidance provided in the memorandum “Updated Guidance for Area Designations for the 2010 Primary Sulfur Dioxide National

Ambient Air Quality Standards”, dated March 20, 2015 and the SO 2 NAAQS Designations Modeling TAD. Indiana used a multi-nested rectangular receptor grid with appropriate spacing of receptors based on the distance from the modeled emission points to detect significant concentration gradients. The modeling domain extended out to include all sources and the

• Receptor spacing at the fence line for each facility was placed every 50 meters. • Receptor spacing at 100 meters was placed out to a distance of 3,000 meters (3 kilometers) beyond each facility. • Receptor spacing at 250 meters was placed out to a distance of 5,000 meters (5 kilometers) beyond each facility. • Receptor spacing at 500 meters was placed out to a distance of 10,000 meters (10 kilometers) beyond each facility.

Page 65 of 69

Figure 16.4 – Duke – Cayuga Receptor Grid

Duke – Cayuga is largely fenced and has regular security patrols to keep unauthorized people off the property. Since this is the case, receptors were placed along the property line. Duke – Cayuga’s concentrations increase extending out from the property line, indicating that maximum modeled concentrations occur further away from the Duke – Cayuga property.

16.7 Stack Heights The use of actual stack heights rather than relying on Good Engineering Practice (GEP) stack heights when modeling actual emissions was utilized in the analysis per the SO 2 NAAQS Designations Modeling TAD.

16.8 Temporally Varying Seasonal 1-Hour SO 2 Background

Temporally varying seasonal SO 2 background concentrations were developed in accordance with the recommended U.S. EPA guidance for establishment of such background concentrations in

Page 66 of 69

Section 8.2 of 40 CFR Part 51, Appendix W and considered appropriate and representative of the area. The latest three years of SO 2 air quality monitoring data (2012-2014) was used.

Temporally varying seasonal 1-hour SO 2 background concentrations were taken from the Fountain County monitor for 2012 - 2014. The hourly seasonal SO 2 values used for representative background concentrations for the area surrounding Duke - Cayuga are listed below in Table 16.3.

Table 16.3 – Duke – Cayuga 99 th Percentile Temporally Varying

Seasonal SO 2 Background Values (ppb) Hr 1 Hr 2 Hr 3 Hr 4 Hr 5 Hr 6 Hr 7 Hr 8

Winter 7.76 7.52 7.00 6.49 8.00 7.00 6.00 6.51 Spring 7.69 8.00 7.55 8.00 8.00 7.53 7.54 6.56 Summer 4.50 5.00 4.00 3.48 3.42 3.00 3.00 3.00 Fall 6.58 5.62 6.00 5.00 7.56 6.57 7.18 6.55

Hr 9 Hr 10 Hr 11 Hr 12 Hr 13 Hr 14 Hr 15 Hr 16

Winter 8.55 9.60 9.98 9.00 9.00 8.26 7.65 8.30 Spring 8.63 9.00 10.00 8.00 8.63 9.00 9.00 7.64 Summer 6.22 7.24 8.62 8.00 9.00 8.00 6.57 6.60 Fall 6.60 6.63 9.00 8.67 8.00 7.62 9.00 8.68

Hr 17 Hr 18 Hr 19 Hr 20 Hr 21 Hr 22 Hr 23 Hr 24 Winter 6.00 8.42 8.62 11.00 8.00 8.18 8.85 8.00 Spring 8.00 8.00 9.00 8.60 9.00 7.00 8.00 7.38 Summer 6.58 5.56 6.58 5.00 4.00 4.00 6.52 4.00 Fall 8.63 8.14 7.55 7.56 6.48 7.53 8.00 7.53

16.9 SO 2 Emissions Included in the Modeling Analysis 16.9.1 DRR Source: Duke - Cayuga Emissions Duke - Cayuga has two units, Units BLR1 and BLR2 that have continuous emission monitoring

(CEM) data for SO 2 from 2012 - 2014. This hourly CEM data from both units was formatted and used in the 1-hour SO 2 AERMOD model run. Total annual emissions from Duke - Cayuga from 2015 are approximately one-half of the emissions from 2012 through 2014 emissions.

Page 67 of 69

Therefore, modeling the 2012-2014 emissions is conservative in nature. The auxiliary boiler will also be modeled based on the 2014 emissions reporting.

16.9.2 Inventoried SO 2 Sources Included in the Modeling

SO 2 sources from the surrounding area were evaluated to determine if their SO 2 emissions had a potential impact on the air quality surrounding the DRR source, beyond what is captured through background monitoring data. The latest available actual emissions over three years (2012-2014) were used. The following list of sources were included in the AERMOD run to determine overall air quality characteristics. Table 16.4 lists the inventory source to be included in the AERMOD run to determine overall air quality characteristics for the area surrounding Duke - Cayuga.

Table 16.4 – Duke – Cayuga Modeling Source Inventory 2012-2014 SO Emissions Source Source ID Location 2 (tpy) Eli Lilly 165-00009 Vermillion County 1618.8 a Colonial Brick 165-00002 Vermillion County 76.5 b a A short-term emission rate for the three-year (2012-2014) average was modeled for Eli Lilly. b A three-year (2012-2014) annual average was calculated for Colonial Brick. Colonial Brick was shut down in 2016. They still have an active Title V permit on file.

16.10 Modeling Results The 99 th percentile of the 1-hour daily maximum modeled concentrations represents the fourth

to the 1-hour SO 2 standard to indicate whether a modeled violation of the SO 2 NAAQS occurred. All concentrations fell below the 1-hour SO 2 NAAQS and were determined to attain the standard and the area surrounding Duke - Cayuga is recommended as attainment. The maximum th predicted 99 percentile daily 1-hour SO 2 concentration is shown in Table 16.5. The overall maximum concentration was 176.4 µg/m 3, occurring at UTM coordinates 458750.0 East, 4421750.0 North.

Table 16.5 – Duke – Cayuga Modeling Results Total Modeled Concentration 1-Hour SO Facility Including Seasonal Hourly 2 Emission Scenarios NAAQS Models Background (µg/m 3) Attainment (µg/m 3) Duke - Cayuga 176.4 196.2 Yes Page 68 of 69

th The concentration isopleths showing the maximum predicted 99 percentile daily 1-hour SO 2 concentration gradients can be found in Figure 16.5.

Figure 16.5 – Duke - Cayuga Modeling Results

Page 69 of 69

ENCLOSURE 1

1-Hour SO 2 Background Determination

U.S. EPA revised the SO 2 National Ambient Air Quality Standard (NAAQS) by instituting a 1- hour primary standard of 75 parts per billion (ppb). Therefore, an analysis was necessary to determine ambient 1-hour SO 2 background concentrations representative for all regions in the state. This determination is needed in order to make attainment designations, attainment demonstrations and perform New Source Review (NSR) and Prevention of Significant

Deteoriation (PSD) modeling. Indiana has reviewed the 1-hour SO 2 monitoring and meteorological data from 2012 through 2014 to calculate representative ambient 1-hour SO 2 background concentrations. U.S. EPA’s “SO 2 NAAQS Designations Modeling Technical Assistance Document, December 2013” was followed to calculate the background concentrations in order to eliminate overly conservative cumulative impacts from nearby major SO 2 emission sources when performing air quality dispersion modeling. Overview

th Indiana has 21 SO 2 monitors located throughout the state. Table 1 shows the 99 percentile for the years 2012, 2013, 2014, and 2015 and the 2012-2014 and 2013-2015 1-hour SO 2 design values for the 7 SO 2 monitors that the attainment designation are based on.

Table 1 - 1-Hour SO 2 Design Values for SO 2 Monitors (ppb) in Indiana 99 th Percentile 2012-2014 2013-2015 Design Design County Monitor ID 2012 2013 2014 2015 Value Value Floyd 18-043-1004 32.0 20.5 43.8 26.0 32 30 Fountain 18-045-0001 30.0 34.0 22.0 19.0 29 25 Jasper 18-073-0002 33.0 40.0 18.0 10.0 30 23 Lake 18-089-0022 47.0 43.2 53.1 35.0 48 44 Porter 18-127-0011 36.0 36.0 27.0 39.0 33 34 Vanderburgh 18-163-0021 16.5 18.6 32.3 18.0 22 23 Vigo 18-167-0018 72.5 79.1 85.0 71.0 79 78

Data Retrieval

Monitoring data for the SO 2 monitors near the DRR sources were retrieved from U.S. EPA’s AirData database. The concentration data were supplied for each hour and day of every month from 2012 through 2014. Meteorological data was collected in order to correlate the wind

Page 1 of 3

directions and concentrations for each hour of each day of every month. Meteorological data was either collected at a monitor near the monitoring site or the nearest National Weather Service (NWS) station or Automated Surface Observation Stations (ASOS). This data was collected and distributed by the Midwest Regional Climate Center (mrcc.isws.illinois.edu). The nearest meteorological data to each of the SO 2 monitors is summarized below.

Table 2 - Locations of SO 2 Monitors and Meteorological Stations for Background Analysis Monitor Station County/Site Monitor ID Location Meteorological Station Location Floyd Co. / 38.31 o N Charlestown State Park 38.39 o N 18-043-1004 New Albany 85.83 o W meteorological station 85.66 o W Fountain Co. / 39.96 o N 39.79 o N 18-045-0001 Indianapolis NWS station North of S.R. 234 87.42o W 86.18 o W Jasper Co. / 41.19 o N 41.69 o N 18-073-0002 South Bend NWS station Wheatfield 87.05 o W 86.25 o W Lake Co. / 41.72 o N Gary IITRI 41.61 o N 18-089-0022 Gary - IITRI 86.91 o W meteorological station 87.30 o W Porter Co. / 41.63 o N Gary IITRI 41.61 o N 18-127-0011 Dunes Acres 87.10 o W meteorological station 87.30 o W Vanderburgh Co. / 38.01 o N 38.05 o N 18-063-0021 Evansville NWS station Buena Vista 87.58 o W 87.52 o W Vigo Co. / 39.49 o N 39.79 o N 18-167-0018 Indianapolis NWS station Lafayette Ave 87.40 o W 86.18 o W

Methodology for Determining Ambient SO 2 Background Concentrations

Each set of SO 2 data was paired with the corresponding meteorological conditions for every hour of the year in order to determine the wind direction for each hour that SO 2 concentrations were recorded. Data was processed in chronological order with daily and seasonal trends analyzed.

The initial analysis created pollution roses to determine the wind directions from which the highest SO 2 concentrations were coming. This analysis helped to identify the nearest upwind SO 2 emission sources impacting the SO 2 monitor. With those wind directions identified, SO 2 concentrations (10 ppb and above) resulting from SO2 emission sources from those wind directions were removed from the analysis, in order to calculate a representative ambient SO 2 background concentration for each SO 2 monitor. This analysis helps to prevent double-counting SO 2 emission source impacts in an air quality modeling analysis. Once data for the SO 2 monitors were processed, the data was re-formatted in order to calculate the hourly-seasonal 99 th percentile averages over a 3-year period, as detailed in U.S. EPA’s “SO 2 NAAQS Designations Page 2 of 3

Modeling Technical Assistance Document, December 2013 Section 8 – Background Concentrations”. The 99 th percentile concentrations, based on each hour of the day and each of the four seasons of the year, were calculated for each SO 2 monitor.

In order to calculate the seasonal hourly 99 th percentile average, the data was grouped by the seasonal months. Spring was represented by concentrations recorded in March, April and May; summer represented by June, July and August; fall represented by September, October and November and winter represented by December, January and February. Once this data was grouped by seasons, the 99 th percentile was calculated for each hour of the day, making 24 th th separate 99 percentiles for each SO 2 monitoring site per season. The average of these 99 percentiles over the three-year period represents the hourly-seasonal 1-hour SO 2 background. Summary

For purposes of the modeling analysis related to the DRR, adjusted 1-hour SO 2 background values were used for the Posey, Floyd, Sullivan, Vermillion, Jasper, Lake and Porter counties

DRR sources. Calculations to determine adjusted 1-hour SO 2 background concentrations were made according to U.S. EPA’s “SO 2 NAAQS Designations Modeling Technical Assistance Document, February 2016 Section 8 – Background Concentrations”. This approach calls for the removal of SO 2 concentrations emitted from large SO 2 emission sources located directly upwind of a SO 2 monitor. This allows for more representative ambient background values to be determined, not overly conservative values that could possibly double-count direct SO 2 source impacts and 1-hour SO 2 background concentrations when modeling inventory sources.

Page 3 of 3

This page left intentionally blank. Enclosure 2 Lake County DRR Source Modeling Inventory Point Sources

Stack Stack Exit Stack SO2 Emission Determination Company Source ID Source Description East (X) North (Y) Height Temperature Velocity Diameter Emissions CEM/Varying/Annual (m) (m) (m) (K) (m/s) (m) (tpy)

1 AMUSA 7 Sinter Plant East Windbox 463341 4612705 48.768 383.15 25.146 3.6576 355.17 CEM

2 AMUSA 26 4SP HMD South 464015 4613844 7.9248 314.26 22.443439 2.2555 3.89 CEM

3 AMUSA 27 4SP HMD North 464049 4613882 5.7912 314.26 22.23516 2.4384 3.89 CEM

4 AMUSA 37 4SP Secondary Vent 464129 4613916 6.096 299.82 15.23492 4.8646 0.95 CEM

5 AMUSA 38 4SP Steelmaking Off Gas 464111 4613786 45.72 338.71 22.9616 3.9624 27.3 CEM

6 AMUSA 101 101 464125 4612000 38.4048 519.26 18.81 1.676 4.71E-04 CEM

7 AMUSA 102 102 464115 4611990 38.4 519.26 18.81 1. 676 4.71E-04 CEM

8 AMUSA 107 107 464100 4612030 67.06 672.04 7.596 3 .3528 0.001052852 CEM

9 AMUSA 108 108 464090 4611930 67.06 672.04 7.596 3 .353 0.001035227 CEM

10 AMUSA 134 5 BH 501-503 464897 4614738 68.58 407.04 14.1224 5.1816 338.15 CEM

11 AMUSA 141 EAF Melting 461960 4610940 43.5864 377.04 2.86512 10.2443 85.93 CEM

12 AMUSA 143 EAF LMF 461859 4610982 13.8684 340.37 18.39976 1.143 13.94 CEM

13 AMUSA 147 2SP 10 Furnace Off Gas 463272 4612185 77.724 1922.04 13.49758 1.8288 28.02 CEM

14 AMUSA 148 2SP 20 Furnace Off Gas 463383 4612297 73.152 1922.04 13.49758 1.8288 28.01 CEM

15 AMUSA 149 2SP Secondary Vent 463461 4612335 64.008 302.04 8.712199 3.6576 11.2 CEM

16 AMUSA 152 2SP HMD 463393 4612307 4.572 316.48 12.79144 3.81 0

17 AMUSA 154 2SP LMF 463202 4612155 18.288 339.82 10.24128 1.8288 20.01 CEM

18 AMUSA 166 IH7 Casthouse Baghouse 2 (W) 464670 4614630 4.572 310.93 33.67531 2.987 203.9 CEM

19 AMUSA 167 IH7 Casthouse Baghouse 1 (E) 464870 4614500 46.9392 327.59 16.03756 3.3528 203.9 CEM

20 AMUSA 170 IH7 Stoves 464800 4614500 70.104 533.15 14.1732 5.1816 398.77 CEM

21 AMUSA 195 IH7 BFG Flare 464870 4614490 55.7784 922.04 2 2.6518 136.9 CEM

1 of 13

Lake County DRR Source Modeling Inventory Point Sources

22 Cokenergy 201 Cokenergy 465354 4614325 89.9 422.04 20.33016 5.4864 5236 CEM

23 IHCC 220 Boiler 504 464920 4614849 96.012 404.97 16.1544 3.048 236.25 CEM

24 AMUSA 45A No. 1 Lime Kiln Bghse Stack A 463894 4613596 21.3055 477.59 18.5674 0.9662 5.55 CEM

25 AMUSA 45B No. 1 Lime Kiln Bghse Stack B 463897 4613600 21.3055 477.59 18.5674 0.9662 5.55 CEM

26 AMUSA 45C No. 2 Lime Kiln Bghse Stack A 463883 4613607 21.3055 477.59 18.5674 0.9662 5.55 CEM

27 AMUSA 45D No. 2 Lime Kiln Bghse Stack B 463887 4613610 21.3055 477.59 18.5674 0.9662 5.55 CEM

28 IHCC IHCCCH1 Charging-Battery A/B 465174 4614512 18.2911 394.26 17.61134 2.7402 2.385 3-yr ave annual

29 IHCC IHCCCH2 Charging-Battery C/D 465150 4614134 18.2911 394.26 17.61134 2.7402 2.385 3-yr ave annual

30 IHCC IHCCPS Pushing 465154 4614232 7.7602 394.26 25.99944 0.8595 6.2 3-yr ave annual

31 IHCC IHCCQ1 Quenching A/B 465264 4614353 18.3002 373.15 3.191256 11.9786 1.95 3-yr ave annual

32 IHCC IHCCQ2 Quenching C/D 465258 4614315 18.3002 373.15 3.191256 11.9786 1.95 3-yr ave annual

33 IHCC IHCCVS IHCC Vent Stacks 465166 4614224.5 25.3 983.15 18.37944 2.3896 2419.7 CEM

34 IHCC102 465199.13 4614569.39 25.2984 983 12.246864 2.3866 0

35 IHCC103 465178.47 4614116.45 25.2984 983 12.246864 2.3866 0

36 IHCC104 465174.04 4614010.86 25.2984 983 12.246864 2.3866 0

37 IHCC105 465202.47 4614661.37 25.2984 983 12.246864 2.3866 0

38 IHCC106 465179.9 4614182.35 25.2984 983 12.246864 2.3866 0

39 IHCC107 465192.44 4614485.78 25.2984 983 12.246864 2.3866 0

40 IHCC108 465189.1 4614403.84 25.2984 983 12.246864 2.3866 0

41 IHCC109 465127.22 4614295.14 25.2984 983 12.246864 2.3866 0

42 IHCC110 465126.54 4614212.11 25.2984 983 12.246864 2.3866 0

43 IHCC111 465123.81 4614104.15 25.2984 983 12.246864 2.3866 0

2 of 13

Enclosure 2 Lake County DRR Source Modeling Inventory Point Sources

44 IHCC112 465114.24 4614024.89 25.2984 983 12.246864 2.3866 0

45 IHCC113 465150.63 4614659.69 25.2984 983 12.246864 2.3866 0

46 IHCC114 465143.95 4614574.41 25.2984 983 12.246864 2.3866 0

47 IHCC115 465138.93 4614495.81 25.2984 983 12.246864 2.3866 0

48 IHCC116 465133.91 4614407.18 25.2984 983 12.246864 2.3866 0

49 AMIH S1A IH3 Stoves 462621 4612774 65.2272 533.15 9.99744 3.2918 105.45 CEM

50 AMIH S1B IH4 Casthouse Baghouse 462629 4612930 22.7076 339.26 8.966201 3.6881 117.03 CEM

51 AMIH S1C IH4 Stoves 462629 4612787 62.1792 533.15 9.99744 3.9929 244.3 CEM

52 AMIH S1D IH4 Bleeder 462645 4612785 31.0896 922.04 5.916168 1.7242 138.7 CEM

53 AMIH S1E IH3 Bleeder 462624 4612765 31.0896 922.04 3.837432 1.7242 81.7 CEM

54 AMIH S301 IH7 Granulator - Lafarge 464750 4614550 99.44 336 5.479999 3.96 28.5 CEM

55 AMIH S3B 3SP HMD Baghouse 462734 4613566 8.8087 304.82 8.102599 1.204 54.65 CEM

56 AMIH S4A HSM Reheat Furnace 1 462645 4614319 65.2272 977.59 8.74776 4.572 0

57 AMIH S4B HSM Reheat Furnace 2 462668 4614311 65.2272 977.59 8.74776 4.572 0

58 AMIH S4C HSM Reheat Furnace 3 462691 4614305 65.2272 977.59 8.74776 4.572 0

59 AMIH S8E No. 6 Boiler 462286 4612566 46.9392 683.15 26.79192 3.048 180.5 CEM

60 AMIH S8G No. 8 Boiler 462273 4612540 46.9392 688.71 17.31264 3.5052 356.7 CEM

61 Ironside S8H No. 9 Boiler 462269 4612577 46.9392 683.15 26.79192 3.048 204.3 CEM

84 US Steel 94011 Sinter Plant Windbox 473218 4607057 56.388 385.93 20.23872 3.4442 534.9 3-yr ave annual

85 US Steel 940541 TBBH Boiler 1 472661 4607149 45.72 572.04 14.478 3.6576 72.18 3-yr ave annual

86 US Steel 940542 TBBH Boiler 2 472661 4607136 45.72 572.04 14.478 3.6576 124.14 3-yr ave annual

3 of 13

Enclosure 2 Lake County DRR Source Modeling Inventory Point Sources

87 US Steel 940543 TBBH Boiler 3 472661 4607123 45.72 572.04 14.478 3.6576 126.2 3-yr ave annual

88 US Steel 940545 TBBH Boiler 5 472661 4607096 45.72 572.04 14.478 3.6576 63 3-yr ave annual

89 US Steel 94053 TBBH Boiler 6 472655 4607079 45.72 499.82 12.16152 3.6576 72.3 3-yr ave annual

90 US Steel 94017 84 inch Hot Strip Mill Reheat Furnaces 468755 4608468 49.6824 701.48 50.81016 2.4689 107.8 Seasonal Varying

91 US Steel 940121 No. 4 BH Boiler 1 472592 4607817 35.3568 460.93 18.83664 2.8956 153.3 Seasonal Varying

92 US Steel 940122 No. 4 BH Boiler 2 472592 4607792 35.3568 460.93 18.83664 2.8956 168.81 Seasonal Varying

93 US Steel 940123 No. 4 BH Boiler 3 472592 4607767 35.3568 460.93 18.83664 2.8956 110.92 Seasonal Varying

94 US Steel 940401 CPBH Boiler 8 474393 4606802 94.1832 535.93 5.66928 3.048 23.6 Seasonal Varying

95 US Steel 940402 CPBH Boiler 9 474436 4606850 60.96 535.93 5.66928 2.8042 23.6 Seasonal Varying

96 US Steel 940403 CPBH Boiler 10 474436 4606866 60.96 535.93 5.66928 2.8042 23.6 Seasonal Varying

97 US Steel 94070 Tail Gas Incinerator 474470 4606815 97.536 894.26 22.86 0.5791 1.2 3-yr ave annual

98 US Steel 94026 No. 2 Underfiring 473903 4606522 106.68 368.71 3.2004 6.096 67.3 3-yr ave annual

99 US Steel 94038 CPBH Boiler 6 474362 4606775 40.5384 535.93 5.334 2.5908 23.6 Seasonal Varying

100 US Steel 94037 CPBH Boilers 4 an 5 474337 4606775 40.5384 535.93 5.334 2.5908 23.6 Seasonal Varying

101 US Steel 94066 No. 14 BF Casthouse 472643 4607841 50.292 329.82 20.4216 3.9624 719.9 Seasonal Varying

102 US Steel 94039 Coke Plant Boiler No. 7 474370 4606803 32.004 535.93 5.12064 2.5908 23.6 3-yr ave annual

103 US Steel 94036 Coke Plant Boiler No. 3 474315 4606782 39.3192 535.93 9.26592 1.8898 23.6 3-yr ave annual

104 US Steel 94021 No. 4 BF Stoves 472694 4606861 68.58 314.82 3.47472 3.9014 53.9 Seasonal Varying

105 US Steel 94022 No. 6 BF Stoves 472697 4607006 68.58 319.82 8.13816 3.9014 92.2 Seasonal Varying

106 US Steel 94023 No. 8 BF Stoves 472701 4607166 76.2 313.71 5.88264 3.9014 59.4 Seasonal Varying

107 US Steel 94013BFSTOVE #13 BF Stoves 472696 4607680 76.2 325.37 6.21792 15.5143 101.7 Seasonal Varying

4 of 13

Enclosure 2 Lake County DRR Source Modeling Inventory Point Sources

108 US Steel 94041 No. 1 BOP Desulf Caster 472325 4606631 24.384 299.82 22.82952 3.109 41.6 3-yr ave annual

109 US Steel 94007 Sinter Cooler 473194 4607100 30.48 455.37 18.8976 5.4864 86.3 3-yr ave annual Precarbon #2 (by Coke Battery #2) includes 110 US Steel USPRECA CASP C 473933 4606552 49.9872 499.98 9.99744 2.0117 3 3-yr ave annual

111 US Steel USBFGFL BFG Flare Stacks (closer to BF #4) 472724 4606895 200.0098 922.04 9.99744 4.9987 90.3 3-yr ave annual

112 US Steel 94045QBOP2 No 2 QBOP Desulf Caster 472524 4607641 16.764 331.48 16.3068 1.1582 0

113 US Steel 940CB5 Coke Battery #5 Underfire 473200 4606400 76.2 499.82 4.38912 3.048 23.8 3-yr ave annual

114 US Steel 940CB7 Coke Battery #7 Underfire 473200 4606600 76.2 533.15 5.6388 3.048 33.1 3-yr ave annual

115 US Steel COGBYPROD Coke Oven Gas Recovery 473200 4606600 30.48 366.48 2.98704 1.0058 0

116 US Steel 940CASPC CASP C 474393 4606802 16.764 366.48 3.048 2.0117 10.36 3-yr ave annual

117 US Steel USCOGFLARE COG stack Desulf 473534.18 4606500.83 45.72 922.04 3.048 5.7912 69.6 3-yr ave annual

118 US Steel US1BOPCAST 472477 4607429 24.4145 394.26 20.20824 2.4384 0

132 BP AMOCO BP1 3SPS Boiler 1 459991.4 4613228.4 18.3948 508.36 7.7852016 0.8083 15.91 3-yr ave annual

133 BP AMOCO BP2 3SPS Boiler 2 459991.4 4613237.1 18.3948 508.36 7.7852016 0.8083 15.76 3-yr ave annual

134 BP AMOCO BP3 3SPS Boiler 3 459973.9 4613228.4 18.3948 508.36 7.7852016 0.8083 15.51 3-yr ave annual

135 BP AMOCO BP4 3SPS Boiler 4 459973.9 4613237.1 18.3948 508.36 7.7852016 0.8083 16.68 3-yr ave annual

136 BP AMOCO BP5 3SPS Boiler 6 459955.8 4613231.5 18.3948 508.36 7.7852016 0.8083 17.45 3-yr ave annual

137 BP AMOCO BP6 FCU 500 CAT 460103 4612576 23.2258 567 10.451592 0.8361 25.1 3-yr ave annual

138 BP AMOCO BP7 11 PS - H-1X 459829.8 4613338.8 18.4877 496.94 2.3783544 0.8826 6.06 3-yr ave annual

139 BP AMOCO BP9 11 PS - H-3 459861.8 4613314.7 15.329 607.12 3.5396424 0.3995 1.25 3-yr ave annual

140 BP AMOCO BP10 11 PS - H-200 459792.5 4613422.1 18.209 536.13 3.0751272 0.9104 7.38 3-yr ave annual

141 BP AMOCO BP11 11 PS- H-300 459792.5 4613443.9 18.209 536.13 2.4990552 0.8268 4.75 3-yr ave annual

5 of 13

Enclosure 2 Lake County DRR Source Modeling Inventory Point Sources

142 BP AMOCO BP12 #1 CRU/ ARU PROCESS HEATER 459703 4612854 16.258 514.53 0.4273296 1.1241 2.13 3-yr ave annual

143 BP AMOCO BP13 #2 CRU/ ARU PROCESS HEATER 459666 4612852 15.6077 631.81 3.0470856 0.4831 1.1 3-yr ave annual

144 BP AMOCO BP14 FCU 600 CAT 459945 4612578 14.8645 605.58 6.1222128 0.7432 16.25 3-yr ave annual

145 BP AMOCO BP15 ALKY 460095 4612741.85 18.1161 962.68 6.094476 0.3066 1.55 3-yr ave annual

146 BP AMOCO BP16 DDU or South ? 800-04 459855.11 4613618.24 18.5806 962.68 6.094476 0.3809 90.97 3-yr ave annual

147 BP AMOCO BP17 FCU 459721.53 4612637.23 18.5806 962.68 6.094476 0.3716 1.25 3-yr ave annual

148 BP AMOCO BP18 4UF 459550 4612830 18.5806 962.68 6.094476 0.6039 13.75 3-yr ave annual

149 BP AMOCO BP19 UIU 459751.57 4612755.58 19.9742 962.68 6.094476 0.4274 7.4 3-yr ave annual

150 BP AMOCO BP20 VRU 460280 4612423.82 18.1161 962.68 6.094476 0.1951 2.39 3-yr ave annual

151 BP AMOCO BP21 ARU - F200A, F-200B 459993 4613060 18.5806 474.41 1.161288 1.0684 8.22 3-yr ave annual

152 BP AMOCO BP22 4UF - F-1, F-8A, F-8B 459707 4613011 15.9793 554.66 2.5270968 1.1241 4.64 3-yr ave annual

153 BP AMOCO BP23 4UF - F-2 459635 4613011 19.7883 548.79 1.9510248 1.0684 3.86 3-yr ave annual

154 BP AMOCO BP24 4UF - F-3 459645 4613011 18.3948 560.52 2.1646896 0.9755 4.14 3-yr ave annual

155 BP AMOCO BP25 4UF - F-4, F-5, F-6 459665 4613011 17.1871 505.27 1.8022824 1.0684 4.31 3-yr ave annual

156 BP AMOCO BP27 New 12 PS Atmospheric Heater H-101A 460629 4612809.3 18.3019 505.27 2.4804624 0.9941 21.4 3-yr ave annual

157 BP AMOCO BP28 New 12 PS Vacuum Heater H-102 460619.9 4612706.6 18.3948 496.01 2.3411688 0.9941 7.78 3-yr ave annual

158 BP AMOCO BP30 New Coker Heater 460567 4612560 18.4877 506.81 2.1089112 0.6968 6.31 3-yr ave annual

159 BP AMOCO BP31 New Coker Heater 460566 4612515 18.4877 506.81 2.1089112 0.6968 6.54 3-yr ave annual

160 BP AMOCO BP32 New Coker Heater 460566 4612477 18.4877 506.81 2.1089112 0.6968 6 3-yr ave annual

161 BP AMOCO BP33 New Hydrogen Plant 461343 4612750 9.2903 505.27 4.645152 1.0498 0

162 BP AMOCO BP34 New Hydrogen Plant 461401 4612695 9.2903 505.27 4.645152 1.0498 0

6 of 13

Enclosure 2 Lake County DRR Source Modeling Inventory Point Sources

163 BP AMOCO BP35 COT1 and COT2 460224 4612806 11.6129 573.17 6.670548 0.576 52.52 3-yr ave annual

164 BP AMOCO BP36 New GOHT Heater 459477.4 4613541.6 13.0064 628.73 4.3665648 0.3345 0

165 BP AMOCO BP37 New 12 PS Atmospheric Heater H-101B 460629 4612839.8 18.3019 505.27 2.4804624 0.9941 0

166 BP AMOCO BP39 ISOM - H-1 459822 4612853 11.6129 517.62 2.6849832 0.7618 4.69 3-yr ave annual

167 BP AMOCO BP42 DDU - WB-301 and WB-301 459443 4613297 13.1922 644.16 5.9457336 0.3623 4.92 3-yr ave annual

168 BP AMOCO BP43 HU - B-501 459586 4613330 23.2258 505.27 5.8527696 0.641 2.23 3-yr ave annual

169 Carmeuse KILN1A Carmeuse 1 466117.95 4610027.09 24.3596 477.59 3.048 1.9812 8.76 SO2 Limit

170 Carmeuse KILN1B 466119.57 4610029.22 24.3596 477.59 3.048 1.9812 8.76 SO2 Limit

171 Carmeuse KILN1C 466121.19 4610031.35 24.3596 477.59 3.048 1.9812 8.76 SO2 Limit

172 Carmeuse KILN1D 466122.81 4610033.47 24.3596 477.59 3.048 1.9812 8.76 SO2 Limit

173 Carmeuse KILN1E 466124.43 4610035.6 24.3596 477.59 3.048 1.9812 8.76 SO2 Limit

174 Carmeuse KILN1F 466126.05 4610037.73 24.3596 477.59 3.048 1.9812 8.76 SO2 Limit

175 Carmeuse KILN2A Carmeuse 2 466108.24 4610034.44 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

176 Carmeuse KILN2B 466109.85 4610036.58 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

177 Carmeuse KILN2C 466111.47 4610038.72 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

178 Carmeuse KILN2D 466113.09 4610040.86 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

180 Carmeuse KILN2F 466116.32 4610045.14 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

181 Carmeuse KILN3A Carmeuse 3 466096.38 4610042.8 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

182 Carmeuse KILN3B 466097.99 4610044.93 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

183 Carmeuse KILN3C 466099.6 4610047.07 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

184 Carmeuse KILN3D 466101.22 4610049.2 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

7 of 13

Lake County DRR Source Modeling Inventory Point Sources

185 Carmeuse KILN3E 466102.83 4610051.34 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

186 Carmeuse KILN3F 466104.44 4610053.47 26.4932 477.59 3.048 1.9812 8.76 SO2 Limit

187 Carmeuse KILN4A Carmeuse 4 466086.05 4610050.06 28.956 477.59 3.048 1.9812 8.76 SO2 Limit

188 Carmeuse KILN4B 466087.66 4610052.19 28.956 477.59 3.048 1.9812 8.76 SO2 Limit

189 Carmeuse KILN4C 466089.27 4610054.33 28.956 477.59 3.048 1.9812 8.76 SO2 Limit

190 Carmeuse KILN4D 466090.88 4610056.46 28.956 477.59 3.048 1.9812 8.76 SO2 Limit

191 Carmeuse KILN4E 466092.49 4610058.6 28.956 477.59 3.048 1.9812 8.76 SO2 Limit

192 Carmeuse KILN4F 466094.1 4610060.73 28.956 477.59 3.048 1.9812 8.76 SO2 Limit

193 Carmeuse KILN5A Carmeuse 5 466076.28 4610057.34 26.8224 477.59 3.048 1.9812 8.76 SO2 Limit

194 Carmeuse KILN5B 466077.89 4610059.47 26.8224 477.59 3.048 1.9812 8.76 SO2 Limit

195 Carmeuse KILN5C 466079.51 4610061.61 26.8224 477.59 3.048 1.9812 8.76 SO2 Limit

196 Carmeuse KILN5D 466081.13 4610063.75 26.8224 477.59 3.048 1.9812 8.76 SO2 Limit

197 Carmeuse KILN5E 466082.74 4610065.88 26.8224 477.59 3.048 1.9812 8.76 SO2 Limit

198 Carmeuse KILN5F 466084.36 4610068.02 26.8224 477.59 3.048 1.9812 8.76 SO2 Limit

199 Koppers KOPPER24 437771.4 4630123.7 25.908 508.1 22.06752 1.524 569.8 3-yr ave annual

200 Koppers KOPPER77 437763 4630123.6 25.908 508.1 22.06752 1.524 569.8 3-yr ave annual

201 Koppers KOPPER53 437576.5 4630111.5 23.4696 794.2 14.23 0.76 333.6 3-yr ave annual

202 Koppers KOPPER76 437577.2 4630093.9 23.4696 777.6 9.31 0.76 312.5 3-yr ave annual

203 AMBH AMSRC12 Battery 2 Pushing Stack 488266.6 4609400.9 64.008 1088.71 41.57472 1.524 0

204 AMBH AM57 BOF Hot Metal Desulf 1 Baghouse Stack 488498.4 4609914.4 25.9111 305.37 12.94892 2.0513 0

205 AMBH AM59 BOF Hot Metal Desulf 2 Baghouse Stack 488512 4609940.1 25.9111 305.37 5.887721 3.0389 0

206 AMBH AM60 BOF Hot Metal Desulf 3 Baghouse Stack 488514.6 4609952.1 12.192 319.26 12.94892 2.664 0

8 of 13

Lake County DRR Source Modeling Inventory Point Sources

207 AMBH P6 AMBurns PwrStn Blr 8-12 488403 4609297 67.9704 505.37 13.939519 3.5113 4312.5 Seasonal Varying

208 NIPSCO PU78FGD NIPSCO Bailly 489738 4610321 146.304 327.59 26.634989 6.2484 1368.7 CEM

211 LaFarge LAFAR1 465166 4614224.5 25.3 983.15 18.37944 2.39 98.45 3-yr ave annual

212 Safety Kleen SK4 460158.59 4610790.08 30.48 1080.37 6.767 1.3716 34.9 3-yr ave annual

213 Safety Kleen SK7 460162.05 4610772.05 30.48 1019.26 4.572 0.9693 21.36 3-yr ave annual

214 Safety Kleen SK8 460153.73 4610772.75 30.48 1055.37 7.132 0.8534 6 3-yr ave annual

215 Eco Service 00242_2 460128.5 4606396.7 10.668 810.93 15.651 1.3716 3.76 3-yr ave annual

216 Eco Service 00242_3 460053.5 4606385.4 91.44 334.26 12.89304 1.8288 251.29 3-yr ave annual

225 AMUSA AMUSA166 2SP BOF Charge Aisle 463400 4612140 4.572 316.48 33.6804 2.987 6.48 3-yr ave annual

226 AMBH P7001 110 Plate Mill #1 & 2 Stack 489029.6 4608811 54.5592 838.71 2.1336 4.4409 0.4 3-yr ave annual

227 AMBH P6503 160 Plate Mill #1 Slab Reheat Furnace 489014 4609043 54.2544 672.04 4.368802 3.1029 15.2 3-yr ave annual

228 AMBH P6504 160 Plate Mill #2 Slab Reheat Furnace 489035 4609043 54.2544 672.04 4.08432 3.2095 16.6 3-yr ave annual

229 AMBH P6509 160 PM #5 IN/OUT REHEAT FURNACE 489053.9 4609039 39.9288 783.15 12.476479 1.9507 0

230 AMBH P6502 160 PM #7 IN/OUT REHEAT FURNACE 489042.2 4608914 32.9184 783.15 9.987281 2.2372 0

231 AMBH P6505 160 PM #8 BATCH FURNACE 489042.2 4608894 50.9016 672.04 2.98704 1.7374 0

232 AMBH P3018 BATTERY #1 PECS 488053.3 4608389 30.48 360.93 25.26585 2.4384 53.61 3-yr ave annual

233 AMBH P3026 #1 Underfire Coke Oven 487967.9 4608346 76.8096 560.93 9.144 3.7795 1759.97 Seasonal Varying

234 AMBH P3024 BATTERY #2 PECS 488059.1 4608115 26.8224 360.93 25.26585 2.4384 60.7 3-yr ave annual

235 AMBH P3027 #2 Underfire Coke Oven 487958.6 4608191 75.8952 560.93 9.144 4.0447 2261.91 Seasonal Varying

236 AMBH P3547 C Furnace Stoves/Stacks (4 stoves) 488244.3 4609339 61.2648 533.15 15.8496 3.4839 864.44 Seasonal Varying

237 AMBH P3560 D Furnace Stoves/Stacks (4 stoves) 488229.2 4609496 61.2648 533.15 14.894558 3.5936 1629.1 Seasonal Varying

238 AMBH P90A HOT STRIP MILL #1 WALKING BEAM FCE E 489029.2 4609235 96.012 810.93 7.061201 3.2004 21.8 3-yr ave annual

9 of 13

Lake County DRR Source Modeling Inventory Point Sources

239 AMBH P90B HOT STRIP MILL #1 WALKING BEAM FCE W 489009 4609235 96.012 810.93 7.061201 3.2004 21.8 3-yr ave annual

240 AMBH P91A HOT STRIP MILL #2 WALKING BEAM FCE E 489051.1 4609236 96.012 810.93 7.02564 3.2004 11.3 3-yr ave annual

241 AMBH P91B HOT STRIP MILL #2 WALKING BEAM FCE W 489030.1 4609235 96.012 810.93 7.02564 3.2004 11.3 3-yr ave annual

242 AMBH P92A HOT STRIP MILL #3 REHEAT FURNACE STACK E 489069 4609236 41.4528 810.93 8.8392 3.9624 27.2 3-yr ave annual

243 AMBH P92B HOT STRIP MILL #3 REHEAT FURNACE STACK W 489053.1 4609236 41.4528 810.93 8.8392 3.9624 27.2 3-yr ave annual

254 AMBH P2501 Power Station Boiler #7 488405.1 4609255 67.9704 505.37 14.43228 3.2004 879.84 Seasonal Varying

255 AMBH P3513 SINTER PLANT WINDBOX SCRUBBER STACK 488038.3 4609329 24.0792 322.04 13.9446 5.1816 702.78 Seasonal Varying

256 AMBH P4002 STEELMAKING HMD STATION #1 488512.1 4609936 25.9111 305.37 12.948919 2.0513 10.7 3-yr ave annual

257 AMBH P59 STEELMAKING HMD STATION #2 488512 4609940 25.9111 305.37 5.887721 3.0389 10.7 3-yr ave annual

260 AMBH P4008 STEELMAKING HMD STATION #3 488514.6 4609952 12.192 319.26 12.948919 2.664 9.6 3-yr ave annual

261 AMBH P3091 Coke Oven Export Gas Flare 487988 4608372 30.48 1922.04 9.397999 0.9144 1.8 3-yr ave annual

262 AMBH P3540 C Furnace BFG Flare (2 flareheads) 488274.8 4609359 64.008 1088.71 41.57472 1.524 18.6 3-yr ave annual

263 AMBH P3553 D Furnace BFG Flare (2 flareheads) 488278.3 4609495 64.008 1088.71 41.57472 1.524 18.64 3-yr ave annual AMUSA - ArcelorMittal - USA Cokenergy - Cokenergy, Inc AMIH - ArcelorMittal - Indiana Harbor Ironside - Ironside Energy, Inc US Steel - U.S. Steel - Gary Works BP AMOCO - BP Products - North America Inc. Carmeuse - Carmeuse Lime, Inc Koppers - Koppers Inc - Illinois AMBH - ArcelorMittal - Burns Harbor NIPSCO - NIPSCO Bailly Generating Station LaFarge - ISPAT Inland LaFarge North America Safety Kleen - Safety Kleen Eco Service - Eco Services Corp (formerly Rhodia, Solvay)

10 of 13

Lake County DRR Source Modeling Inventory Volume Sources

Release Initial Horizontal Initial Vertical Emission Determination Company Source ID Source Description East (X) North (Y) Height Dimension Dimension SO2 Emissions CEM/Varying/Annual (m) (m) (m) (m) (m) (tpy)

62 AMIH V3B1 3SP HMD Fugitives 462672 4613541 16.15 2.23 7.51 0.278 3-yr ave annual

63 AMIH V3B2 3SP HMD Fugitives 462734 4613566 16.15 2.23 7.51 0.278 3-yr ave annual

64 AMIH V3B3 3SP HMD Fugitives 462717 4613529 16.15 2.23 7.51 0.278 3-yr ave annual

65 AMIH V3B4 3SP HMD Fugitives 462738 4613525 16.15 2.23 7.51 0.278 3-yr ave annual

66 AMIH V1A1 IH3 Casthouse 462562 4612734 23.8 0.85 11.1 1.41 3-yr ave annual

67 AMIH V1A2 IH3 Casthouse 462561 4612733 23.8 0.85 11.1 1.41 3-yr ave annual

68 AMIH V1A3 IH3 Casthouse 462560 4612731 23.8 0.85 11.1 1.41 3-yr ave annual

69 AMIH V1A4 IH3 Casthouse 462559 4612730 23.8 0.85 11.1 1.41 3-yr ave annual

70 AMIH V1A5 IH3 Casthouse 462558 4612728 23.8 0.85 11.1 1.41 3-yr ave annual

71 AMIH V1A6 IH3 Casthouse 462556 4612727 23.8 0.85 11.1 1.41 3-yr ave annual

72 AMIH V1A7 IH3 Casthouse 462555 4612725 23.8 0.85 11.1 1.41 3-yr ave annual

73 AMIH V1A8 IH3 Casthouse 462554 4612724 23.8 0.85 11.1 1.41 3-yr ave annual

74 AMIH V1A9 IH3 Casthouse 462553 4612722 23.8 0.85 11.1 1.41 3-yr ave annual

75 AMIH V1A10 IH3 Casthouse 462552 4612721 23.8 0.85 11.1 1.41 3-yr ave annual

76 AMIH V1B1 IH4 Casthouse 462697 4612866 31.1 0.99 14.5 3.27 3-yr ave annual

77 AMIH V1B2 IH4 Casthouse 462696 4612864 31.1 0.99 14.5 3.27 3-yr ave annual

78 AMIH V1B3 IH4 Casthouse 462695 4612863 31.1 0.99 14.5 3.27 3-yr ave annual

79 AMIH V1B4 IH4 Casthouse 462693 4612861 31.1 0.99 14.5 3.27 3-yr ave annual

80 AMIH V1B5 IH4 Casthouse 462692 4612860 31.1 0.99 14.5 3.27 3-yr ave annual

81 AMIH V1B6 IH4 Casthouse 462690 4612858 31.1 0.99 14.5 3.27 3-yr ave annual

82 AMIH V1B7 IH4 Casthouse 462689 4612856 31.1 0.99 14.5 3.27 3-yr ave annual

11 of 13

Lake County DRR Source Modeling Inventory Volume Sources

83 AMIH V1B8 IH4 Casthouse 462688 4612855 31.1 0.99 14.5 3.27 3-yr ave annual

119 US Steel 221 CB2UNDERFUG 473900 4606300 19.9949 9.9974 9.9974 10.8

120 US Steel 222 CB5UNDERFUG 473913 4606438 19.9949 9.9974 9.9974 1

121 US Steel 447110 #4 BF Casthouse Roof M 472679.5 4606687.4 18.1051 4.2977 8.7996 6.32 Emission Factor

122 US Steel 447210 #4 BF Casthouse Roof M 472685.4 4606667.7 18.1051 4.2977 8.7996 6.32 Emission Factor

123 US Steel 447310 #4 BF Casthouse Roof M 472691.3 4606648 18.1051 4.2977 8.7996 6.32 Emission Factor

124 US Steel 447410 #6 BF Casthouse Roof M 472683 4606848 17.4986 4.2977 8.7996 6.34 Emission Factor

125 US Steel 447510 #6 BF Casthouse Roof M 472688.9 4606828.3 17.4986 4.2977 8.7996 6.34 Emission Factor

126 US Steel 447610 #6 BF Casthouse Roof M 472694.7 4606808.5 17.4986 4.2977 8.7996 6.34 Emission Factor

127 US Steel 447710 #8 BF Casthouse Roof M 472686.7 4606991.9 17.1999 4.2977 8.3972 5.9 Emission Factor

128 US Steel 447810 #8 BF Casthouse Roof M 472692.5 4606972.2 17.1999 4.2977 8.3972 5.9 Emission Factor

129 US Steel 447910 #8 BF Casthouse Roof M 472698.4 4606952.4 17.1999 4.2977 8.3972 5.9 Emission Factor

130 US Steel 448110 #13 BF Casthouse RM 472710.6 4607478.3 34.3997 6.3978 15.999 12.75 Emission Factor

131 US Steel 448210 #13 BF Casthouse RM 472713.1 4607461.2 34.3997 6.3978 15.999 12.75 Emission Factor

209 AMBH P133 488222 4609449 50 16 3.6576 0

210 AMBH P134 488220 4609591 50 16 3.6576 0

217 AMIH AMIH142 461896 4610979 47.5488 4.9378 21.97 0

218 AMIH AMIH165 464750 4614615 21.9456 7.4981 1.4204 0

219 F1C 462531 4612706 49.9994 15.999 12 37.5 3-yr ave annual

220 F1D 462726 4612870 49.9872 15.999 12 69.5 3-yr ave annual

221 AMUSA 171A IH7 Casthouse Fugitives 464721 4614598 21.9456 7.6352 19.56 0

222 AMUSA 171B IH7 Casthouse Fugitives 464731 4614598 21.9456 7.6352 19.56 0

12 of 13

Lake County DRR Source Modeling Inventory Volume Sources

223 AMUSA 171C IH7 Casthouse Fugitives 464741 4614598 21.9456 7.6352 19.559 0

224 AMUSA 171D IH7 Casthouse Fugitives 464751 4614598 21.9456 7.6352 19.559 0

244 AMBH PFE101 488022.5 4608137.9 16.43 13.6 7.65 0.465817359 3-yr ave annual

245 AMBH PFE102 488023.4 4608163.5 16.43 13.6 7.65 0.465817359 3-yr ave annual

246 AMBH PFE103 488022.8 4608185.1 16.43 13.6 7.65 0.465817359 3-yr ave annual

247 AMBH PFE104 488023.1 4608208.7 16.43 13.6 7.65 0.465817359 3-yr ave annual

248 AMBH PFE105 488024.3 4608231.3 16.43 13.6 7.65 0.465817359 3-yr ave annual

249 AMBH PFE201 488012.9 4608305.6 16.43 13.6 7.65 0.465817359 3-yr ave annual

250 AMBH PFE202 488013.2 4608327.3 16.43 13.6 7.65 0.465817359 3-yr ave annual

251 AMBH PFE203 488012.7 4608349.1 16.43 13.6 7.65 0.465817359 3-yr ave annual

252 AMBH PFE204 488013.1 4608375.5 16.43 13.6 7.65 0.465817359 3-yr ave annual

253 AMBH PFE205 488013.9 4608397.5 16.43 13.6 7.65 0.465817359 3-yr ave annual

258 AMBH BFDCHFUG 488240 4609560 24.7 21.4 3.5 14.5 3072061 3-yr ave annual

259 AMBH BFCCHFUG 488242 4609426 24.7 21.4 3.5 14.5 3072061 3-yr ave annual

13 of 13

This page left intentionally blank. ENCLOSURE 3 Carmeuse Commissioner’s Order

1 of 8

2 of 8

3 of 8

4 of 8

5 of 8

6 of 8

7 of 8

8 of 8

ENCLOSURE 4 SABIC Commissioner’s Order

1 of 6

2 of 6

3 of 6

4 of 6

5 of 6

6 of 6

Attachment 3

U.S. Mineral Products (Isolatek) Discussion This page left intentionally blank. U.S. Mineral Products (Isolatek - Source ID: 069-00021)

The Indiana Department of Environmental Management (IDEM) excluded U.S. Mineral Products (USM) d/b/a Isolatek International, a mineral wool manufacturer near Huntington, Indiana in Huntington County, from its January 7, 2016 list of affected sources to be characterized under the Data Requirements Rule (DRR). Per the thresholds established within the DRR, USM’s most current reported sulfur dioxide (SO 2) emissions were well below levels required for the rule to be applicable. However, United States Environmental Protection Agency (U.S. EPA) identified USM as an additional source to be characterized in its March 25, 2016 response to IDEM.

Indiana strongly objects to the inclusion of USM as an affected source under the DRR. The

DRR defines applicable sources as stationary sources that had actual SO 2 emissions in 2014 of 2,000 tons or more, or have been identified by IDEM or U.S. EPA “as requiring further air quality characterization.” (40 CFR § 51.1202). Indiana did not include USM on its list of

sources subject to the DRR because its reported actual SO 2 emissions in 2014 were 164 tons, less than one tenth of the DRR threshold of 2,000 tons or more. According to U.S. EPA’s calculations (based on an informal in-house 2007 stack test), USM’s actual annual emissions would have been “approximately 444 tons of SO 2” in 2014. U.S. EPA also determined that 2014 was an abnormally low year for production and estimated 800 tons of SO 2 per year during normal production years, which is still less than half the DRR emission threshold. USM has seen a downturn since 2013 in its wool production (approximately 40,000 tons/year could be considered a prior normal), with a slight bounce back to 28,000 to 30,000 tons per year production over the last few years. This is still much lower than historic production, but should be considered the current normal production at the facility based upon current economic factors with the economy.

USM has operated the same equipment at its Huntington facility since 1982. In its March 25,

2016, letter, U.S. EPA indicated an emission factor of 21.6 lb SO 2 per ton of melt was appropriate for the USM cupola emissions. USM has historically used an emission factor of 8 lbs/ton based upon U.S. EPA’s Compilation of Air Pollutant Emission Factors, AP-42. 1 As a result of a Clean Air Act (CAA) §114 information request, USM submitted to U.S. EPA, a summary sheet from stack tests previously conducted which included some engineering studies from 2007 and several pages from the 2007 study report for in-house testing of particulate matter

(PM), nitrogen oxides (NO x) and SO 2 at the facility. That study included an informational emission test for SO 2 for the cupola that was only performed in the downdraft ducts. The results were reported in the summary sheet and in the study report. USM does not consider 21.6 lbs/ton

1 An emission factor of 0.2 lbs/hr was used from 2000 through 2005 as a result of an error in the data used for the annual emission calculation. When the correction was made in 2005, IDEM advised USM that it was not necessary to correct the prior emission statements. Page 1 of 5

to be a valid SO 2 emission factor due to problems with the cupola operation at the time of the informal test. It should be noted that IDEM did not review or approve of an SO 2 stack test protocol in 2007 for USM and had no compliance inspector present at the informal SO 2 test. According to production records available for the time period on and around the stack testing days in December 2007, the following may be concluded as summarized by USM:

On Dec 17 th , the first day of the testing, both cupolas were idled in the morning for a period exceeding 2 hours each due to an electrical problem with a charge hoist. In addition, #1 cupola idled for 3 hours directly preceding the hoist issue due to a spinner motor failure. Typically, following an idle period of time, the cupola operating conditions take some time (could easily be several hours) to stabilize. Thus, the testing period started with less than normal conditions.

th On Dec 18 , during the period of the testing for SO 2 data collection, #1 cupola went through a period of increased coke consumption and reduced melt rate. Both indicators were showing a variation from standard coke consumption and melt rate in the 10% - 20% order of magnitude. USM standard coke consumption is expected to be at ~320 – 340 lbs / ton of charge and the avg. melt rate at ~4.2 tons / hr. At the time of the stack test USM recorded an avg. of 360 – 380 lbs of coke / charge and a melt rate of ~3.9 tons / hr respectively. These variations are considered significant and clearly not normal operations. Those conditions are related and indicate that the operator was attempting to overcome the slower melt rate by adding additional coke to the charge. Based upon the increased coke consumption and slower melt rate, general operating conditions at the time of testing are best described as poor. The raw material receivers from that period of time indicate a higher than normal moisture content in received coke (10%-15% vs. standard of <7%) explaining the need for additional BTUs with every charge to evaporate the excess moisture. The low moisture content of coke is a critical factor to the cupola performance. During the period of time in 2007 around when the testing was performed, the USM coke supplier was struggling to provide a product with acceptable quality. The coke quality issues were caused by operational issues at the source. USM had no viable, alternative supply options at the time.

In order to resolve the emission factor issue, in 2016, USM conducted an engineering study of the cupola emissions. This consisted of an informational emission test that included SO 2 measured in the baghouse. That test indicated an SO 2 emission factor range of 9.22 to 9.36 lbs/ton. The results of the 2016 test confirm that the emission factor from AP-42 is appropriate to use for the USM mineral wool cupola operation. Additionally, the 164 tons of SO 2 reported as actual emissions for USM should be considered valid for DRR purposes. This is significantly lower than what U.S. EPA is attempting to rely upon in its analysis.

U.S. EPA identified the 2,000 ton threshold as an important indicator of the need for prioritized air quality characterization under the DRR. U.S. EPA set the threshold at a level “that prioritizes

the resources that will be devoted to characterizing air quality near SO 2 sources nationally.” (80 Page 2 of 5

FR 51061). That threshold is already on “the lower end of the range of thresholds” of sources that have the potential to contribute to violations of the National Ambient Air Quality Standard (NAAQS) (80 FR 51061). Furthermore, that threshold “strikes a reasonable balance between the need to characterize air quality near sources that have a higher likelihood of contributing to a NAAQS violation and the analytical burden on air agencies.” (80 FR 51061). U.S. EPA did not characterize the 2,000 ton threshold as an arbitrary number, but rather as an indicator of sources warranting prioritization of state and federal resources.

Because USM’s actual SO 2 emissions and total potential-to-emit SO 2 emissions remain well below the 2,000 ton applicability threshold, it is unreasonable to place it among the sources that should be prioritized to determine if it contributes to violations of the NAAQS. Including sources with actual SO 2 emissions of less than one-tenth the 2,000 ton threshold represents a misapplication of the intent of the DRR to prioritize sources and resources. Indiana believes that this reinterpretation of the DRR inappropriately broadens the scope and purpose of this phase of the DRR. There are numerous sources across the United States that fall into a similar category as USM. In Indiana alone, there are thirty five (35) sources with reported actual emissions between that of USM and the 2,000 ton threshold. Among these is a manufacturer of mineral wool, with very similar operational characteristics, with reported actual emissions greater than that of USM, and sources located in densely populated areas with as much as ten times the reported emissions of USM, which happens to be located in a sparsely populated rural area. Based on familiarity with how the dispersion model handles certain operations, it is safe to assume that some of these sources would clearly pose a greater threat to the NAAQS and human health than USM. Therefore, U.S. EPA’s identification of USM is clearly arbitrary and capricious.

Due to the time constraints that U.S. EPA has placed on states to implement the DRR, broadening the applicability of the DRR’s phased approach thwarts the rule’s intent to prioritize state and federal resources. IDEM does not question whether the DRR provides states or U.S. EPA the authority to identify sources with actual emissions below the 2,000 ton threshold as requiring further air quality characterization. However, if this is done, it should be done consistently and not arbitrarily. U.S. EPA did not use a systematic approach to identify sources below 2,000 tons that have the greatest probability to pose a risk to exceeding the NAAQS and threaten human health. Therefore, IDEM disagrees that USM should be arbitrarily subjected to further characterization under the DRR.

Page 3 of 5

Table 1: U.S. Mineral Products (Isolatek) 2016 Stack Test Data

Page 4 of 5

Page 5 of 5

This page left intentionally blank. Attachment 4

Alcoa Warrick Attainment Discussion This page left intentionally blank.

ALCOA Warrick Power Plant (Source ID: 173-00007) and Warrick Operations (Source ID: 173-00002)

Aluminum Manufacturing Company of America (ALCOA) operates an aluminum manufacturing facility and power plant in Newburgh, Indiana, along the northern bank of the Ohio River in Anderson Township, Warrick County. The Indiana Department of Environmental Management (IDEM) believes Warrick Power Plant and Warrick Operations and the surrounding area should be designated attainment based on historical sulfur dioxide (SO 2) ambient monitoring data showing attainment of the SO 2 standard.

Warrick Power and Warrick Operations were determined to be sources subject to United States Environmental Protection Agency’s Data Requirements Rule (DRR) based on actual 2014 SO 2 emissions of 4,993 tons and 3,500 tons, respectively. However, Warrick Operations shut down its smelter operations as of March 31, 2016, and has therefore ceased to generate potline point SO 2 emissions, potline smelter line source SO 2 emissions, or SO 2 emissions from the anode baking ring furnace. Warrick Operations currently operates a rolling mill that uses natural gas and will generate SO 2 emissions of less than one ton per year.

Historical SO 2 data from monitors operated by ALCOA prior to the shut-down of the smelter show attainment of the 2010 primary SO 2 1-hour National Ambient Air Quality Standard (NAAQS). Tables 1 and 2 provide data from SO 2 monitors operated by ALCOA for several years prior to the smelting operation’s shutdown. The tables also show data from the SO 2 monitor that IDEM operates in the region near the ALCOA Warrick facility.

As shown in the tables, all 99 th percentile values since 2009 are below the 1-hour standard. In addition, the most recently available design value (2008 – 2010), and all recent partial-year design values, are also below the 1-hour standard. These low monitor values occurred during the time when the operations plant and the power plant were in full operation. As such, it is reasonable to conclude that the area surrounding the ALCOA Warrick facilities is in attainment of the 1-hour SO 2 NAAQS. As such, Indiana is recommending Anderson Township, Warrick County, Indiana as attainment.

Page 1 of 2

th Table 1: Warrick County SO 2 Monitor Data - 99 Percentile Values (parts per billion) (2005 – 2016)

Site ID 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 IDEM Operated 181630021 66 67 69 41 17 18.8 19.4 16.5 18.6 32.3 18 11 a ALCOA Operated 181730002 143 199 103 111 38 18 Not Operational 23 36 b 181730004 Not Operational 63 57 b 181730005 Not Operational 46 42 b 181730012 Not Operational 59 62 b a – Data through July 31, 2016. b – Data through June 30, 2016.

Table 2: Warrick County SO 2 Monitor Data – Design Values (parts per billion) (2007 – 2016)

2005 - 2006 - 2007 - 2008 - 2009 - 2010 - 2011 - 2012 - 2013 – 2014 – Site ID 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 IDEM Operated 181630021 67 59 43 26 18 18 18 22 23 21 ALCOA Operated 181730002 148 138 84 56 Not Operational 23 a 30 b a b 181730004 Not Operational 63 60 a b 181730005 Not Operational 46 44 a b 181730012 Not Operational 59 61 a – Based on one year of data. b – Based on two years of data.

Page 2 of 2

Attachment 5

ArcelorMittal – Burns Harbor SO2 Air Quality Monitor System Documentation This page left intentionally blank. U.S. EPA Confirmation Letter for Siting Methodologies of

ArcelorMittal Burns Harbor SO2 Monitor

Page 1 of 22

Page 2 of 22

U.S. EPA’s Approval of IDEM’s 2017 Ambient Air Monitoring Network Plan

Page 3 of 22

Page 4 of 22

Clean Air Engineering Certifications for

ArcelorMittal Burns Harbor’s SO2 Monitor

Page 5 of 22

Page 6 of 22

Page 7 of 22

Page 8 of 22

Page 9 of 22

Page 10 of 22

Page 11 of 22

Page 12 of 22

Page 13 of 22

Page 14 of 22

Page 15 of 22

Page 16 of 22

Page 17 of 22

Page 18 of 22

Page 19 of 22

Page 20 of 22

Page 21 of 22

Page 22 of 22